**3. Diagnosis of endometritis**

The diagnosis of endometritis in the mare represents one of the crucial challenges for the equine clinical practice [55]. Equine practitioners have recently understood that the major cause of the progression from an acute to a chronic condition is the failure to identify the causative agent, given that affected mares do not always show evident clinical signs [55]. Hence, it is critical to make a correct etiological diagnosis, and to characterize the degree and the type of inflamma‐ tion, in order to establish an effective treatment. Furthermore, it is not uncommon that mares with a long history of normal fertility can acquire post-breeding endometritis. In such cases, the clinician has no opportunity for prophylactic intervention [56].

The diagnostic algorithm for the identification of the causative agent comprises a detailed reproductive history and a complete clinical evaluation, including the transrectal palpation of the reproductive tract and the use of ancillary diagnostic aids [17, 57–58]. Among the latters, specific instrumental investigations, such as transrectal ultrasonography, vaginal and cervical exams, with both manual and endoscopic evaluation, and laboratory tools, such as uterine culture, endometrial cytology and biopsy, represent useful tools for the diagnosis of endome‐ tritis [17, 58]. Recently, myeloperoxidase has been investigated as a uterine inflammatory biomarker, but the relationship between its concentration and pathologic uterine conditions needs further studies [59]. To increase the pregnancy rates in "problem" mares, the practitioner should dispose a quick and reliable diagnostic technique to start the best treatment as soon as possible in the breeding season [56].

#### **3. 1. Experimental identification of susceptible mares**

Mares with endometritis begin to show intrauterine fluid accumulation following breeding, after the first three to four successful pregnancies. Uterine degenerative changes appear and worsen with age, the reproductive tract may become more pendulous and uterotonic drugs are no longer effective [57]. Furthermore, some reproductively normal, nulliparous as well as pluriparous mares exhibit an excessive uterine inflammatory response after being bred with frozen semen. The exuberant reaction may be a consequence of a low volume of seminal plasma and the lack of its beneficial action in decreasing the migration of neutrophils into the uterine lumen [57].

The delay in uterine clearance has an important role in the pathogenesis of uterine inflamma‐ tion, since it influences the retention of endometrial inflammatory factors, whether they are bacterial or sperm-induced [60]. In fact, it was shown that mares susceptible to chronic uterine infections failed to clear adequately 51Cr-labeled microspheres when compared with mares with normal uteri [37, 60]. The uterus, studied with scintigraphy in mares with a delayed uterine clearance, showed to be oriented vertically, while in reproductively normal mares it was oriented horizontally. Hence the "baggy" uterus contributes to fluid accumulation and low clearance [37, 57]. However, although scintigraphy has been proven to be a useful method for studying post-breeding endometritis, it seems not to be a functional diagnostic tool in a clinical setting [61]. In addition to the uterine position, the myometrial function plays an important role in uterine clearance, as confirmed by electromyographic recordings [60]. Nevertheless, not all mares susceptible to endometritis exhibit a delayed uterine clearance if cervical dilatation is appropriate [25].

#### **3. 2. Clinical identification of susceptible mares**

#### *3. 2. 1. History*

**2. 2. Bacterial endometritis**

290 Genital Infections and Infertility

manipulations [48–49].

treatment failure or antibiotic resistance [17].

**2. 3. Chronic endometritis**

**3. Diagnosis of endometritis**

The sexually transmissible disease in horses are caused by primary pathogens *Taylorella equigenitalis*, certain unspecified serotypes of *Pseudomonas aeruginosa* and *Klebsiella pneumo‐ niae* capsule types 1, 2 and 5 [45]. The true venereal disease is caused by *T. equigenitalis* and is known as contagious equine metritis resulting in cervicitis, vaginitis and endometritis. Asymptomatic infected carrier stallions transmit the pathogens to the mares, which will present copious muco-purulent vaginal discharge within a week after breeding. However, this pathology is endemic in Europe and a more insidious form with minimal clinical signs has been recognized [46–47]. *P. aeruginosa* and *K. pneumoniae* inhabit the external genital of the stallion and infections are transmitted by coitus, insemination with infected semen and genital

Contamination of the uterus by fecal and genital opportunistic flora may provoke an infection during mating or genital tract manipulations. A wide variety of opportunistic aerobic and anaerobic bacteria, fungi and yeasts, either alone or in synergy, have been occasionally implicated as causes of endometritis. In a report, *Streptococcus equi* subsp. *zooepidemicus* was responsible for approximately 65% of the cases while *Escherichia coli, K. pneumoniae* and *P. aeruginosa* accounted for approximately 10% [47]. The alteration of uterine defense mechanisms is attributed to individual microbial factors such as induction of inflammation, epithelial adherence, resistance to phagocytosis and viscosity of secretions [50]. Bacterial products can modify any properties of mucus, rendering cilia unable to expel uterine exudates. Changes in production, viscosity or elasticity of mucus and cilia function determinate the adverse effects on uterine clearance and hence, these can interfere with antibiotic penetration, resulting in

Chronic infection and mixed population of microorganism cause severe, progressive and irreversible fibrotic condition that affects mare endometrium [51–54]. Long-standing influx of lymphocytes and plasmacells into the endometrium contribute to chronic degenerative changes, such as periglandular, perivascular or diffuse stromal fibrosis. This condition impairs

The diagnosis of endometritis in the mare represents one of the crucial challenges for the equine clinical practice [55]. Equine practitioners have recently understood that the major cause of the progression from an acute to a chronic condition is the failure to identify the causative agent, given that affected mares do not always show evident clinical signs [55]. Hence, it is critical to make a correct etiological diagnosis, and to characterize the degree and the type of inflamma‐ tion, in order to establish an effective treatment. Furthermore, it is not uncommon that mares with a long history of normal fertility can acquire post-breeding endometritis. In such cases,

endometrial function and future pregnancies, causing infertility [52–53].

the clinician has no opportunity for prophylactic intervention [56].

Mares susceptible to developing a persistent post-breeding endometritis are more often pluriparous, barren, older than 12 years of age and with a poor perineal conformation [57, 62]. Nonetheless, some maiden mares, which show an insufficient degree of relaxation of their cervix during estrus, may also develop persistent mating-induced endometritis, regardless of age. In such mares, the problem may be resolved with the first foal delivery [57].

#### *3. 2. 2. Clinical examination*

The mares' reproductive tract examination should include a body condition scoring and the notation of previous or existing foot problems [57]. Any changes in perineal, vulvar or cervical conformation should be evaluated before breeding [56–57, 63]. The two vulvar labia may not create a proper seal due to persistent relaxation of the vulva, and the rectum may be displaced cranially, stretching the vulva dorsally to the ischiatic arch and causing repeated fecal contamination of vulva and vestibule [56]. This kind of conformation is linked to the loss of fat in the perineal area due to a high level of fitness, and muscular fatigue and estrus exacerbate the perineal relaxation, but it can also be a heritable trait [63–64]. It is very important to check that the cervix opens properly in estrus and closes in diestrus [56].

#### **3. 3. Clinical signs**

Clinical signs of endometritis may be hidden, but vaginal discharge, short inter-estrus intervals and/or a shortened luteal phase and reduced fertility can be detected [17, 37, 65]. In bacterial endometritis, they can vary widely according to the pathogen, with symptoms markedly different between Gram-positive and Gram-negative bacteria [17]. A vaginal discharge may be seen more often during estrus [57]. However, the presence of evident clinical signs 48 hours post-breeding is of concern, since a non-inflammatory uterine environment is important for embryonic survival [16].

#### **3. 4. Imaging of the reproductive system: ultrasonography**

Ultrasonography (US) has always been considered an accurate and essential step in the diagnostic workup of endometritis [56, 59, 66]. The monitoring function of US is as important as the diagnostic one, since mares with subclinical endometritis require careful daily checking post-breeding [17]. Since its inception, the use of US has strongly impacted on the ability to detect the presence and to quantify the volume of uterine fluid accumulation, which is considered the main sign of endometritis, especially when observed after breeding; but this "hallmark" sign is not always present [17, 60, 67]. Intrauterine fluid presence has also been shown to be strongly related with low pregnancy rates, especially in mares older than 16 years, and to the biopsy score, revealing its diagnostic value in identifying the uterine inflammation [62, 67–68]. It may be detected as soon as 6 to 36 hours after breeding, but it can be noted at the pregnancy examination 14 to 16 days after breeding [57, 67]. Nonetheless, the more significant time to detect the uterine fluid collection is mid-to-late diestrus because the cervix is closed, and in a healthy uterus, fluid should be little or absent [68]. Intrauterine fluid has been also correlated to specific endometrial bacteria and cytological findings, with mares positive to non-pathogenic bacteria showing intra-uterine fluid in <40% of the US examinations [66–67]. The relevance of the uterine fluid accumulation can be evaluated by examining the depth, the echo-texture and persistence of the fluid [16]. Two or more centimeters of intrauterine fluid during estrus, or between 6 and 36 hours post-breeding, are significant indices of susceptibility to mating-induced endometritis [62, 67]. A high echogenicity in the uterine fluid is associated with the presence of neutrophils and debris, thus indicative of inflammatory components [16]. Short, thick and hyperechoic lines within the uterus may represent air or exudates [17]. In Figure 1, some characteristic appearances of the equine uterus are depicted.

cervix during estrus, may also develop persistent mating-induced endometritis, regardless of

The mares' reproductive tract examination should include a body condition scoring and the notation of previous or existing foot problems [57]. Any changes in perineal, vulvar or cervical conformation should be evaluated before breeding [56–57, 63]. The two vulvar labia may not create a proper seal due to persistent relaxation of the vulva, and the rectum may be displaced cranially, stretching the vulva dorsally to the ischiatic arch and causing repeated fecal contamination of vulva and vestibule [56]. This kind of conformation is linked to the loss of fat in the perineal area due to a high level of fitness, and muscular fatigue and estrus exacerbate the perineal relaxation, but it can also be a heritable trait [63–64]. It is very important to check

Clinical signs of endometritis may be hidden, but vaginal discharge, short inter-estrus intervals and/or a shortened luteal phase and reduced fertility can be detected [17, 37, 65]. In bacterial endometritis, they can vary widely according to the pathogen, with symptoms markedly different between Gram-positive and Gram-negative bacteria [17]. A vaginal discharge may be seen more often during estrus [57]. However, the presence of evident clinical signs 48 hours post-breeding is of concern, since a non-inflammatory uterine environment is important for

Ultrasonography (US) has always been considered an accurate and essential step in the diagnostic workup of endometritis [56, 59, 66]. The monitoring function of US is as important as the diagnostic one, since mares with subclinical endometritis require careful daily checking post-breeding [17]. Since its inception, the use of US has strongly impacted on the ability to detect the presence and to quantify the volume of uterine fluid accumulation, which is considered the main sign of endometritis, especially when observed after breeding; but this "hallmark" sign is not always present [17, 60, 67]. Intrauterine fluid presence has also been shown to be strongly related with low pregnancy rates, especially in mares older than 16 years, and to the biopsy score, revealing its diagnostic value in identifying the uterine inflammation [62, 67–68]. It may be detected as soon as 6 to 36 hours after breeding, but it can be noted at the pregnancy examination 14 to 16 days after breeding [57, 67]. Nonetheless, the more significant time to detect the uterine fluid collection is mid-to-late diestrus because the cervix is closed, and in a healthy uterus, fluid should be little or absent [68]. Intrauterine fluid has been also correlated to specific endometrial bacteria and cytological findings, with mares positive to non-pathogenic bacteria showing intra-uterine fluid in <40% of the US examinations [66–67]. The relevance of the uterine fluid accumulation can be evaluated by examining the depth, the echo-texture and persistence of the fluid [16]. Two or more centimeters of intrauterine fluid during estrus, or between 6 and 36 hours post-breeding, are significant indices

age. In such mares, the problem may be resolved with the first foal delivery [57].

that the cervix opens properly in estrus and closes in diestrus [56].

**3. 4. Imaging of the reproductive system: ultrasonography**

*3. 2. 2. Clinical examination*

292 Genital Infections and Infertility

**3. 3. Clinical signs**

embryonic survival [16].

**Figure 1.** Trans-rectal ultrasonographic evaluation of the equine uterus with a linear transducer (bandwidth 5–10 MHz). In A: normal appearance of the uterus during early diestrus, 48 hours post-ovulation. In B: mild intrauterine fluid accumulation, with moderate edema of uterine folds during early estrus. In C: moderate intrauterine fluid accu‐ mulation, with severe edema of uterine folds during estrus. In D: severe (over 2 cm in depth) intrauterine fluid accu‐ mulation in diestrus, 13 days after ovulation, "hallmark" of equine endometritis.

Another important US sign of endometritis is abnormal edema [17]. Edema may occur physiologically due to lymphangiectasia and under influence of estrogen, i. e. , during the early estrus. A modified version of the subjective system described by Samper has been developed to score uterine edema: grade 0 coincides with the absence of edema; grade I, in which it is difficult to identify uterine folds; grade II, characterized by some of the endometrial folds detectable and a cervix with a fish-bone aspect; grade III, with endometrial folds, easily identifiable and characterized by hyperechoic borders and hypoechoic centers (cartwheel); and grade IV for mares with "hyperedema", in which thickened and bulging endometrial folds are abnormally thick, with hyperechoic border and marked central hypoechogenicity, and the normal architecture of the cartwheel is lost [59]. Cervical incompetence, abnormal vascular reaction to estrogens, a lymphatic pathology or an altered myoelectric activity are all possible non-inflammatory causes of hyperedema [17, 59]. Thus, some edema patterns, such as excessive edema pre- or post-mating, or edema pattern that does not extend throughout the uterine wall, represent a pathological feature [17]. In particular, hyperedema has been considered the marker of uterine pathology when found during the estrus phase, independ‐ ently from the positivity of a uterine cytology, but not during the early estrus [59]. Sometimes uterine fluid and edema are associated, probably due to the severity of drainage deficiency [59]. Furthermore, the appearance and dynamics of uterine cysts can be studied by US. Their effect on pregnancy rate seems to be a quantitative one, with only severely affected mares showing a reduction of fertility [68].

The use of more advanced US software, like Color-, Power-flow- and Pulse-Wave- Doppler applications, allows the evaluation of uterine vascularization. In humans, abnormal uterine blood flow and higher uterine artery impedance have been observed in women with recurrent pregnancy loss and different causes of infertility. Poor blood flow of the gravid uterus has been correlated with advanced age and diffuse endometrial degenerative changes during early pregnancy in mares. Furthermore, disturbed uterine blood flow has been recently associated to other uterine pathologies such as uterine cysts and endometrial elastosis in subfertile mares [69].

#### **3. 5. Endoscopic evaluation of the reproductive system**

After repeated mating in a season, mares may accumulate intrauterine fluid and display classic signs of inflammation on vaginoscopy. Endoscopic examination is the only way to establish the degree and the clinical significance of superficial intrauterine lesions and it is useful in several pathological conditions. It can be used in the mare suspected of having subclinical endometritis due to focal lesions, intra-uterine adhesions and endometrial cups retention. It has proven to be useful in mares with a history of silent heats, which may display endometrial scarring or loss of endometrial folds [17]. The best time to perform endoscopy is during diestrus or early estrus, since it is easier to perform than in other phases. Before the examination, the uterine lumen is dilated with air or saline, but the best visibility is obtained with air. It is important to remember to discard air out of the uterus after the procedure, with a catheter or a pump, because of possible irritation [17].

#### **3. 6. Sampling techniques for bacteriological, cytological and histological evaluation: uterine swab, cytobrush, low volume flush and endometrial biopsy**

The diagnostic methods used to characterize endometritis are the uterine swab, the cytobrush, low volume flush and the endometrial biopsy [37, 70]. As always, the clinician must interpret the data resulting from such techniques together with the clinical signs and bearing in mind how the results of each technique vary, depending on the pathogen [17, 58, 61, 65, 70–73]. All the procedures described thereafter assume that the mares are restrained in an examination stock, the tail is wrapped in an examination glove and suspended to the stock, and the whole perineal area is cleaned and disinfected to avoid contamination from the environment [16, 58, 71]. Each method has its pro and cons and many comparisons of the different techniques have been indeed performed [58, 65, 70–73]. Histologic evaluation of endometrial biopsies is the "gold standard", to which each technique has been compared to calculate the sensitivity and specificity [17, 65, 71]. Nevertheless, it should be remembered that all techniques can yield false-negative results if not conducted by expert clinician [37].

#### *3. 6. 1. Uterine swab*

ently from the positivity of a uterine cytology, but not during the early estrus [59]. Sometimes uterine fluid and edema are associated, probably due to the severity of drainage deficiency [59]. Furthermore, the appearance and dynamics of uterine cysts can be studied by US. Their effect on pregnancy rate seems to be a quantitative one, with only severely affected mares

The use of more advanced US software, like Color-, Power-flow- and Pulse-Wave- Doppler applications, allows the evaluation of uterine vascularization. In humans, abnormal uterine blood flow and higher uterine artery impedance have been observed in women with recurrent pregnancy loss and different causes of infertility. Poor blood flow of the gravid uterus has been correlated with advanced age and diffuse endometrial degenerative changes during early pregnancy in mares. Furthermore, disturbed uterine blood flow has been recently associated to other uterine pathologies such as uterine cysts and endometrial elastosis in

After repeated mating in a season, mares may accumulate intrauterine fluid and display classic signs of inflammation on vaginoscopy. Endoscopic examination is the only way to establish the degree and the clinical significance of superficial intrauterine lesions and it is useful in several pathological conditions. It can be used in the mare suspected of having subclinical endometritis due to focal lesions, intra-uterine adhesions and endometrial cups retention. It has proven to be useful in mares with a history of silent heats, which may display endometrial scarring or loss of endometrial folds [17]. The best time to perform endoscopy is during diestrus or early estrus, since it is easier to perform than in other phases. Before the examination, the uterine lumen is dilated with air or saline, but the best visibility is obtained with air. It is important to remember to discard air out of the uterus after the procedure, with a catheter or

**3. 6. Sampling techniques for bacteriological, cytological and histological evaluation:**

The diagnostic methods used to characterize endometritis are the uterine swab, the cytobrush, low volume flush and the endometrial biopsy [37, 70]. As always, the clinician must interpret the data resulting from such techniques together with the clinical signs and bearing in mind how the results of each technique vary, depending on the pathogen [17, 58, 61, 65, 70–73]. All the procedures described thereafter assume that the mares are restrained in an examination stock, the tail is wrapped in an examination glove and suspended to the stock, and the whole perineal area is cleaned and disinfected to avoid contamination from the environment [16, 58, 71]. Each method has its pro and cons and many comparisons of the different techniques have been indeed performed [58, 65, 70–73]. Histologic evaluation of endometrial biopsies is the "gold standard", to which each technique has been compared to calculate the sensitivity and specificity [17, 65, 71]. Nevertheless, it should be remembered that all techniques can yield

**uterine swab, cytobrush, low volume flush and endometrial biopsy**

false-negative results if not conducted by expert clinician [37].

showing a reduction of fertility [68].

**3. 5. Endoscopic evaluation of the reproductive system**

a pump, because of possible irritation [17].

subfertile mares [69].

294 Genital Infections and Infertility

The uterine swab collection methods for bacteriologic and cytological analysis are the mainstay for the diagnosis of acute endometritis in the mare [74]. Through this technique, bacteria as well as inflammatory cells can be collected and examined by culturing it, or smearing the swab for cytology [70, 74]. Uterine swabs are the most commonly used procedure because of low costs, ease of collection and safety of use; hence, routine pre-breeding uterine swabs should be always obtained from mares "at risk" [56, 65].

The double-guarded swab technique implicates the use of a double sheath, which allows minimizing vaginal contamination [16, 70–71]. The tip of the double-guarded swab is kept covered and free from lubricant as it is introduced into the reproductive tract. Then, it is advanced through the cervix into the uterus and the inner sheath is pushed through the outer sheath. The examiner starts moving the swab to sample the endometrial surface using a pushing and rolling motion for up to 1 min, redirecting the swab into different areas of the uterus. At the end of the sampling, the swab is pulled within the inner sheath, which is drawn back into the outer sheath, and the entire unit is then removed from the reproductive tract [16, 70–71].

Another kind of guarded uterine instrument is the Knudsen catheter that can be autoclaved and used repeatedly. It consists of a metal tube of 87 cm in length and an inner spiral metal rod, and it has a small hole at the tip, to allow advancing a cotton swab. The tube includes a thickened area, the olive, which marks the point of the catheter that should be placed at the outer cervical orifice [74].

For the classic uterine swab, cytological smears are prepared by gently rolling the side of the swab, and by pushing the end on a sterile slide. The swab is sent to laboratory for microbio‐ logical tests, including the tip in a transport container [16, 70–71]. On the other hand, the Knudsen catheter use provides the cotton swab for bacteriology, whereas the cytological sample is obtained by gently removing the material on the spiral rod [74]. Even if this techni‐ que, when properly adopted, is ideal for bacteriology, it leads to contrasting results for the cytological sampling. The use of this technique decreased strongly the number of mares improperly treated, but it may lead to cells deformation depending on the pressure applied during the smearing procedure [70, 74]. Nevertheless, too little pressure during the rolling procedure may lead to clumping of cells [74]. Moistening and gentle rolling have been proven to relief distortion and fragmentation of the collected cells [70]. Furthermore, this method can yield false negative results. In fact, about half of the cases of infectious endometritis resulted negative, since the microorganisms were located in the most pendulous part of uterus that is not easily reached. Conversely, false positive results can be associated with a contaminated sampling [55]. Furthermore, it collects a small superficial endometrial area, resulting in a small amount of total cells [70].

Otherwise, the degree of cellularity obtainable with the Knudsen catheter is influenced by the level of uterine secretion. Indeed, the contact of the smooth surface of the metallic spiral with a dry endometrial surface provides poor cellularity, whereas in case of abundant secretions, samples obtained show an excellent cellularity due to the larger surface area, without cells distortion and with infrequent red blood cells [74].

#### *3. 6. 2. Cytobrush*

Uterine brushes have been used for uterine cytology in humans, as well as in the equine species [60]. As described for the uterine swabs, a guarded technique is used, and the properly lubricated guarding tube is passed through the vagina to the base of a uterine horn or into the uterine body. Once in the area selected for the sampling, the outer tube is retracted far enough to expose the brush, and the cytobrush is rotated in a clockwise direction while in contact with the uterine wall [70]. The instrument is then retracted into the tube prior to removal from the uterus and then rolled on slides.

Cytobrush has been considered superior to the other methods for cytological sampling, since it is easier, more consistent and produces samples with higher cellularity than other techni‐ ques, but a gentle preparation of smear is mandatory to reduce the cells distortion [70, 74]. Cells fragmentation has been frequently observed in cytobrush smears probably due to the rigid fibers that damage the cells, as well as the common occurrence of red blood cells, which is an index of its invasiveness [70]. Proteinaceous material may contaminate the smears obtained with this technique, as well [70]. Nonetheless, it allows collecting cells both from the surface and from the depth of endometrium, up to glandular cells [70]. Such technique could preferentially be used to detect subclinical endometritis in field practice, for its inexpensive‐ ness and safeness, particularly in comparison to endometrial biopsy [58, 65].

#### *3. 6. 3. Low volume flush*

The low volume flush is a method for uterine fluid collection [55]. It is almost as efficient as endometrial biopsy for microorganisms' isolation and can be performed during estrus or diestrus [55]. The procedure involves the use of a low volume (60 to 150 ml) of phosphatebuffered saline (PBS) or lactated Ringer's solution or physiological saline injected into the uterus. The practitioner, properly gloved, advances a sterile insemination pipette, or a Bivona catheter, *per vaginam* into the uterus, through which the solution is injected. Thus, the uterus is transrectally massaged to distribute the fluid evenly throughout the uterine lumen. Mean‐ while, the pipette is moved back and forth and suction is applied to trap cells in the pipette. Finally, the effluent fluid is recovered in a sterile container. If the mare is in estrus, intravenous administration of 10 IU of oxytocin is suggested, to facilitate the release of fluid trapped in the edematous endometrial folds [16, 55, 70, 72].

The fluid is first evaluated macroscopically, holding the sample up to the light, for cloudiness and amount of mucus. The efflux is graded as clear, cloudy or clear with mucus strains. Both mucus and cloudiness are strongly correlated to the presence of *Streptococcus β haemolyticus* and *E. coli* [55, 72]. The recovered fluid is then centrifuged at 400 rpm for 10 min, the super‐ natant is discarded and the pellet is aliquoted in two parts, one for microbiological culture and the other one is resuspended in 1 ml of PBS and drops of the suspension are spilled onto slides for cytological evaluation [16, 55, 70, 72].

The low volume flush is a suitable method to collect cells over a larger surface area and provide more information about cells, mucus and/or exudates than other techniques [70]. It has been showed to be twice as sensitive as swab culture [71–72]. Indeed, it allows quick identification of Gram-negative bacteria, e. g. , *E. coli* [55, 72]. Even if low volume flush requires a larger equipment to be performed than other procedures, it is considered rapid and accurate, and it may prove to be valuable for subclinical endometritis in the chronically infected mare. Nevertheless, this method may cause irritation of the endometrial mucosa and it may be more likely source of contamination from vaginal flora [60, 72]. Such contamination may yield false positive culture results; hence, other indices of endometritis should be added to improve the diagnostic power, such as a rise in pH or presence of debris [70, 72]. Moreover, the samples may show a high number of blood red cells, probably due to the transrectal manipulation of uterus and the scraping effect of the tip of the catheter [70].

Recently, a double-guarded low volume flush technique has been developed, with improved sensibility and specificity in identifying endometritis in the mare [18]. Overall, this new method represents a valid alternative to the classic low volume flush and seems to decrease the risk of contamination during sampling procedure. Furthermore, the availability of a disposable lavage tube and of a closed fluid-tubing system is favorable to be used on field and allows the execution by only one person [18]. On the other hand, the technique showed a poorer ability to find PMNs compared to biopsy [18].

#### *3. 6. 4. Endometrial biopsy*

*3. 6. 2. Cytobrush*

296 Genital Infections and Infertility

uterus and then rolled on slides.

*3. 6. 3. Low volume flush*

edematous endometrial folds [16, 55, 70, 72].

for cytological evaluation [16, 55, 70, 72].

Uterine brushes have been used for uterine cytology in humans, as well as in the equine species [60]. As described for the uterine swabs, a guarded technique is used, and the properly lubricated guarding tube is passed through the vagina to the base of a uterine horn or into the uterine body. Once in the area selected for the sampling, the outer tube is retracted far enough to expose the brush, and the cytobrush is rotated in a clockwise direction while in contact with the uterine wall [70]. The instrument is then retracted into the tube prior to removal from the

Cytobrush has been considered superior to the other methods for cytological sampling, since it is easier, more consistent and produces samples with higher cellularity than other techni‐ ques, but a gentle preparation of smear is mandatory to reduce the cells distortion [70, 74]. Cells fragmentation has been frequently observed in cytobrush smears probably due to the rigid fibers that damage the cells, as well as the common occurrence of red blood cells, which is an index of its invasiveness [70]. Proteinaceous material may contaminate the smears obtained with this technique, as well [70]. Nonetheless, it allows collecting cells both from the surface and from the depth of endometrium, up to glandular cells [70]. Such technique could preferentially be used to detect subclinical endometritis in field practice, for its inexpensive‐

The low volume flush is a method for uterine fluid collection [55]. It is almost as efficient as endometrial biopsy for microorganisms' isolation and can be performed during estrus or diestrus [55]. The procedure involves the use of a low volume (60 to 150 ml) of phosphatebuffered saline (PBS) or lactated Ringer's solution or physiological saline injected into the uterus. The practitioner, properly gloved, advances a sterile insemination pipette, or a Bivona catheter, *per vaginam* into the uterus, through which the solution is injected. Thus, the uterus is transrectally massaged to distribute the fluid evenly throughout the uterine lumen. Mean‐ while, the pipette is moved back and forth and suction is applied to trap cells in the pipette. Finally, the effluent fluid is recovered in a sterile container. If the mare is in estrus, intravenous administration of 10 IU of oxytocin is suggested, to facilitate the release of fluid trapped in the

The fluid is first evaluated macroscopically, holding the sample up to the light, for cloudiness and amount of mucus. The efflux is graded as clear, cloudy or clear with mucus strains. Both mucus and cloudiness are strongly correlated to the presence of *Streptococcus β haemolyticus* and *E. coli* [55, 72]. The recovered fluid is then centrifuged at 400 rpm for 10 min, the super‐ natant is discarded and the pellet is aliquoted in two parts, one for microbiological culture and the other one is resuspended in 1 ml of PBS and drops of the suspension are spilled onto slides

The low volume flush is a suitable method to collect cells over a larger surface area and provide more information about cells, mucus and/or exudates than other techniques [70]. It has been showed to be twice as sensitive as swab culture [71–72]. Indeed, it allows quick identification

ness and safeness, particularly in comparison to endometrial biopsy [58, 65].

The endometrial biopsy is a very helpful tool for diagnosing the cause of subclinical endome‐ tritis, always relating the microscopic lesions to mares' age and reproductive history [17]. Its results about the changes within the endometrium are considered the most reliable. Indeed, it is used to look for degenerative and inflammatory changes, which are classified using both the histologic grading system proposed by Kenney & Doig and modified by Schoon and coll. and the classification of Ricketts [17, 65]. Endometrial biopsies are collected using a sterilized dedicated biopsy instrument, which is passed through the cervix within the uterine lumen. One arm is then inserted into the rectum to guide the biopsy forceps to the desired location, usually at the base of one uterine horn [58, 71]. Once the forceps is closed, it is withdrawn and the sample is macroscopically evaluated for consistency and size, and immediately put in the selected media, depending on its further use.

Bacteriological culture and cytology from endometrial biopsy were demonstrated to be superior to culture swabs, both for sensitivity and for positive predictive value [71]. Endome‐ trial biopsy is safe, but it is not particularly practical [58]. A practical disadvantage is the time between sampling and histologic results, while bacteriological and cytological results are achievable in a short time [58]. Moreover, it is objectively more invasive than other techniques and needs specific equipment, requires further processing, such as shipping to skilled laboratories, time for examination and transmission of results. Hence, it takes longer to have final diagnosis [65, 71].

#### **3. 7. Endometrial cytology**

Endometrial cytology is an inestimable tool in assessing the endometrial inflammation, mainly through the detection of PMNs. Unfortunately, an agreement on the classification and interpretation of cytological results has not yet been reached and different interpretative cutoffs were proposed [16, 70, 74–75]. Indeed, some authors record the number of PMNs as a percentage of all cells seen on a slide and the cut-off value for positive to endometritis ranges from 0. 5 to 5%, according to the guidelines of Brook [16, 71]. The latter classification system categorizes cytological samples as: non-inflammatory (PMNs < 5%), mild inflammation (PMNs 5–15%), moderate inflammation (PMNs 15–30%) and severe inflammation (PMNs > 30%)[16, 70, 75]. Other authors record the amount of PMNs seen per microscopic high power field (HPF) examined. In details, cytological smears are graded as not inflammatory (0–2 neutrophils/ field), moderate inflammation (2–5 neutrophils/field), severe inflammation (>5 neutrophils/ field) or hypocellular (scant epithelial cells and no neutrophils) [70, 75]. The presence of uterine fluid during estrus has been found to be associated with an increased number of PMNs, and mares with intrauterine fluid on the second–third day of estrus were 1. 4 times more likely to have more than 5 PMNs per HPF than those with no or mild inflammation [66–67]. Recently, it has been suggested to evaluate endometrial cytology using the percentage of PMNs in relation to epithelial cells, rather than counting the number of PMNs per HPF, when using the cytobrush for sampling [73]. Nonetheless, the number of PMNs per HPF was found to be inversely proportional to the pregnancy rate, i. e. , mares without inflammation had pregnancy rates 1. 3 and 3 times higher than those of mares exhibiting moderate or severe inflammation, respectively [75].

The amount of PMNs detectable is affected by various factors. Samples obtained in early estrus may not contain inflammatory cells, since PMNs' migration into the uterine lumen during the period of waning progesterone dominance is lower than during maximal estrogen dominance [70, 75]. Nonetheless, the time post-ovulation does not seem to display any effect on endome‐ trial cytological parameters in non-bred mares. Moreover, a small resident amount of PMNs in the endometrium has been demonstrated from 24 to 96 hours after ovulation as well as during pro-estrus, in healthy mares [16, 65, 70–71]. The PMNs number increases in the uterine stratum compactum, but not in low volume flushes, after infusion of semen extenders, saline or seminal plasma, which are known to have an inflammatory effect [65].

Other important parameters useful to interpret the cytological sample are: the background content of the slides, i. e. , if it is proteinaceous, contaminated with red blood cells, or clear; the quality of the cells harvested, i. e. , if they are intact, distorted, or fragmented; the total cellularity, that is the number of cells per HPF; the ratio between PMNs and uterine epithelial cells; the presence of other inflammatory cells, e. g. , eosinophils and monocytes, and their number per HPF; the presence and the number per HPF of vaginal epithelial cells; and the eventual presence of bacteria [55, 70].

Cytology is at the top among the diagnostic techniques, since it is a relatively inexpensive method to obtain results in a short time [73]. However, this method has a relatively high rate of false negative results and it does not provide information about the cause of the inflamma‐ tion [58, 70]. That is why it should always be conducted together with bacteriology, as the detection of PMNs together with potential pathogens is a stronger indicator of endometritis, and the number of mares identified as positive to endometritis is significantly higher than with either technique alone [58, 65, 70–71]. Furthermore, cytological samples supported by positive cultures show on average twice neutrophils than those associated with negative cultures, regardless of the virulence of the bacteria individuated [74]. Nonetheless, mares may have positive cytology with negative culture, and *vice versa* [55, 65]. Various interpretations have been made to explain the absence of a correlation between the two techniques, in particular to justify negative cultures, such as that uterine swabs may miss focal infections, the presence of antimicrobial preparations in the uterus, deep-seated infection or non-infectious irritation [65].

#### **3. 8. Endometrial microbiology**

interpretation of cytological results has not yet been reached and different interpretative cutoffs were proposed [16, 70, 74–75]. Indeed, some authors record the number of PMNs as a percentage of all cells seen on a slide and the cut-off value for positive to endometritis ranges from 0. 5 to 5%, according to the guidelines of Brook [16, 71]. The latter classification system categorizes cytological samples as: non-inflammatory (PMNs < 5%), mild inflammation (PMNs 5–15%), moderate inflammation (PMNs 15–30%) and severe inflammation (PMNs > 30%)[16, 70, 75]. Other authors record the amount of PMNs seen per microscopic high power field (HPF) examined. In details, cytological smears are graded as not inflammatory (0–2 neutrophils/ field), moderate inflammation (2–5 neutrophils/field), severe inflammation (>5 neutrophils/ field) or hypocellular (scant epithelial cells and no neutrophils) [70, 75]. The presence of uterine fluid during estrus has been found to be associated with an increased number of PMNs, and mares with intrauterine fluid on the second–third day of estrus were 1. 4 times more likely to have more than 5 PMNs per HPF than those with no or mild inflammation [66–67]. Recently, it has been suggested to evaluate endometrial cytology using the percentage of PMNs in relation to epithelial cells, rather than counting the number of PMNs per HPF, when using the cytobrush for sampling [73]. Nonetheless, the number of PMNs per HPF was found to be inversely proportional to the pregnancy rate, i. e. , mares without inflammation had pregnancy rates 1. 3 and 3 times higher than those of mares exhibiting moderate or severe inflammation,

The amount of PMNs detectable is affected by various factors. Samples obtained in early estrus may not contain inflammatory cells, since PMNs' migration into the uterine lumen during the period of waning progesterone dominance is lower than during maximal estrogen dominance [70, 75]. Nonetheless, the time post-ovulation does not seem to display any effect on endome‐ trial cytological parameters in non-bred mares. Moreover, a small resident amount of PMNs in the endometrium has been demonstrated from 24 to 96 hours after ovulation as well as during pro-estrus, in healthy mares [16, 65, 70–71]. The PMNs number increases in the uterine stratum compactum, but not in low volume flushes, after infusion of semen extenders, saline

Other important parameters useful to interpret the cytological sample are: the background content of the slides, i. e. , if it is proteinaceous, contaminated with red blood cells, or clear; the quality of the cells harvested, i. e. , if they are intact, distorted, or fragmented; the total cellularity, that is the number of cells per HPF; the ratio between PMNs and uterine epithelial cells; the presence of other inflammatory cells, e. g. , eosinophils and monocytes, and their number per HPF; the presence and the number per HPF of vaginal epithelial cells; and the

Cytology is at the top among the diagnostic techniques, since it is a relatively inexpensive method to obtain results in a short time [73]. However, this method has a relatively high rate of false negative results and it does not provide information about the cause of the inflamma‐ tion [58, 70]. That is why it should always be conducted together with bacteriology, as the detection of PMNs together with potential pathogens is a stronger indicator of endometritis, and the number of mares identified as positive to endometritis is significantly higher than with either technique alone [58, 65, 70–71]. Furthermore, cytological samples supported by positive cultures show on average twice neutrophils than those associated with negative cultures,

or seminal plasma, which are known to have an inflammatory effect [65].

respectively [75].

298 Genital Infections and Infertility

eventual presence of bacteria [55, 70].

A positive uterine culture before breeding is among the causes of infertility linked to endo‐ metritis [17, 75]. If a mare does not spontaneously eliminate an infection in 2 to 4 days, she is considered either persistently infected or on her way of becoming persistently infected, as supported by the study on uterine clearance of bacteria in healthy mares [60]. Mares are classified as resistant when able to clear intrauterine fluid, inflammatory cells, and bacteria within 48 hours from breeding, otherwise they are considered susceptible [61]. In addition, the mares may change their susceptibility over subsequent breeding seasons, in a gradual manner. Some of them exhibit a decreased endometrial quality, while others show a floating resistance [61]. Most susceptible mares exhibit minimal signs of inflammation prior to the first breeding of the year, likely because of prolonged sexual rest [57]. Although uterine culture may yield false positive or false negative results, aerobic endometrial culture is still the most common method for diagnosing infectious endometritis [65, 71]. Indeed, although bacteria may be recovered indicating an infectious endometritis, the underlying problem may be a persistent post-breeding endometritis, either not managed or treated suitably [57]. Bacteria and other microorganisms (yeast or fungi) may be found on cytological smears, free or phagocytized within neutrophils or macrophages, and should be scored based on their number per HPF [16]. Cultures may be positive for one or more bacterial species, but mixed cultures of more than three pathogens are usually considered as the result of contamination [58, 65, 71].

Various correlations have been demonstrated between endometrial bacteria and different US and endometrial cytological findings. Mares with intrauterine fluid were 1. 4 times more likely to have 5 PMNs per HPF on cytological specimens than those with no or mild fluid [66]. Intrauterine fluid was more commonly detected when *hemolytic Streptococcus*, *Klebsiella* species, *Enterobacter cloacae* or yeast were isolated, compared with *E. coli*, *Staphylococcus aureus* and *Pseudomonas* species [66, 72]. Also intrauterine fluid and a cytology containing 2 PMNs per HPF were more commonly associated to *hemolytic Streptococcus* than *E. coli* [66]. Nevertheless, less than 40% of mares from which were isolated *E. coli*, *S. aureus*, *Pseudomonas* spp. , or nonpathogen bacteria, such as *Micrococcus* spp. , *Alpha Streptococcus* or *Bacillus* spp. , had intrau‐ terine fluid at US, immediately before the uterine culture was done [66]. Intrauterine fluid was seen more frequently, in 45–55% of the US examinations, when *β-hemolytic Streptococcus*, *K. pneumoniae*, *E. cloacae* or yeast were isolated [66]. However, intrauterine fluid, especially during estrus, was not always associated with bacterial endometritis [66]. Furthermore, pathogens associated with uterine fluid were more likely to coincide to neutrophils findings on cytology and vice versa [67]. Hence, the uterine fluid indicates an acute inflammation, but not neces‐ sarily a bacterial infection. This is confirmed by the presence of other causes of acute and neutrophilic intrauterine collection, such as pneumovagina, irritating effect of semen, urine reflux into the uterus and excessive production of endometrial mucus [67]. Moreover, not all microbes cause a neutrophilic response and the amount of cytological specimens graded as positive for inflammation varied among the microbial findings [75]. Indeed, *β-hemolytic Streptococcus* or *Klebsiella* yield more often positive uterine cytology, whereas *E. coli*, *S. aureus* and *Pseudomonas* spp. had fewer positive cytological results [66, 75]. Presence of PMNs was strongly associated with *S. equi* subsp. *zooepidemicus*, rather than *E. coli*[71, 74–75]. Among the main pathogens, *β-hemolytic Streptococcus*, in particular *S. equi* subsp. *zooepidemicus*, a Grampositive bacteria, is more likely to cause an endometrial inflammatory response and it was more often associated with positive cytology, differently from other pathogens [74].

Various bacterial species have proven to be pathogenic in the mares' uterine environment, and each species have been isolated in different percentage. Most frequently isolated bacteria, the relative percentage, the sampling techniques adopted and the geographical area in which the study was conducted are summarized in Table 1. In general, *β-hemolytic Streptococcus* is strongly related to a rise in pH in the low volume flush efflux, probably due to super antigens, released by the bacteria itself, which beckon pro-inflammatory mediators into the uterine lumen [72]. They were reported as the most common microorganisms isolated from mares' uterus [71–72, 75]. The double-guarded low volume flush allowed to isolate more *β-hemolytic Streptococcus*, even if uterine lavage seems to be not proper to detect it, probably due to its deep localization in the endometrium [18]. Furthermore, in some *S. equi* subsp. *zooepidemicus* strains, a clear ability to form persistent cells, highly tolerant to penicillin, has been demonstrated. The prevalence of subclinical *S. zooepidemicus* endometritis in the mare is still to be determined, and in subfertile mares the prevalence has been determined to be as high as 64%. The activation of dormant bacteria, using bacterial growth mediums like *bActivate*, may improve significantly the sensitivity of traditional diagnostics [76].

On the other hand, *E. coli*, a Gram-negative bacteria, was associated neither with cytologi‐ cal evidence of inflammation nor with increased pH [65, 71–72, 74–75]. The absence of correlation between *E. coli* and positive cytology may be due to its inability to attract PMNs into the uterine lumen [18]. In fact, the pathogen–host relationship, and consequently the uterine inflammatory response, of *E. coli* appears to be different from that of *β-hemolytic Streptococcus* [72]. On the other hand, it appears to associate with moderate to heavy debris on cytological smears. The interpretation of cultures positive for *E. coli* is difficult, since only pure cultures and large numbers of colonies are considered significant, and an increased sensitivity seems obtainable by culturing low volume flush [71–72, 74]. *E. coli*is sometimes considered a contaminant, but it is believed to cause subfertility in the mare [65], and it has been associated with focal infections, in the form of granulomatous plaques identified with endoscopic evaluation in two mares [72]. Another pathogen is *P. aeruginosa*, a Gramnegative bacteria, which produces a biofilm, i. e. , an adhesive matrix that harbors bacterial microcolonies and resists as well to antibiotics as to the immune system [67]. *P. aeruginosa* is believed to be cause of chronic and persistent infections [67].

The role of non-pathogenic bacteria isolated from the mares' uterus is still unknown, but they may be the cause of decreased pregnancy rates [75]. Mares with bacteriological positivity but no cytological evidence of PMNs may be affected by contaminants [58, 72, 75]. However, their pregnancy rate resulted lower than those of the mares without any positivity to the tests. For example, *Bacillus* and *Micrococcus* are skin commensals that are not usually the cause of lowered pregnancy rates, but in barren and older mares with weakened physical barriers they may play a role in decreased fertility [75]. Even if *S. zooepidemicus* is a commensal and common bacterium of the caudal reproductive tract, the cervix has been considered an efficient barrier maintaining a sterile uterine environment. Trans-cervical medical procedures, particularly when involving instillation of substances that may possibly favor bacterial growth, represent a risk of iatrogenic contamination of the uterus with bacteria from the micro-flora of the caudal reproductive tract [76].

positive for inflammation varied among the microbial findings [75]. Indeed, *β-hemolytic Streptococcus* or *Klebsiella* yield more often positive uterine cytology, whereas *E. coli*, *S. aureus* and *Pseudomonas* spp. had fewer positive cytological results [66, 75]. Presence of PMNs was strongly associated with *S. equi* subsp. *zooepidemicus*, rather than *E. coli*[71, 74–75]. Among the main pathogens, *β-hemolytic Streptococcus*, in particular *S. equi* subsp. *zooepidemicus*, a Grampositive bacteria, is more likely to cause an endometrial inflammatory response and it was

Various bacterial species have proven to be pathogenic in the mares' uterine environment, and each species have been isolated in different percentage. Most frequently isolated bacteria, the relative percentage, the sampling techniques adopted and the geographical area in which the study was conducted are summarized in Table 1. In general, *β-hemolytic Streptococcus* is strongly related to a rise in pH in the low volume flush efflux, probably due to super antigens, released by the bacteria itself, which beckon pro-inflammatory mediators into the uterine lumen [72]. They were reported as the most common microorganisms isolated from mares' uterus [71–72, 75]. The double-guarded low volume flush allowed to isolate more *β-hemolytic Streptococcus*, even if uterine lavage seems to be not proper to detect it, probably due to its deep localization in the endometrium [18]. Furthermore, in some *S. equi* subsp. *zooepidemicus* strains, a clear ability to form persistent cells, highly tolerant to penicillin, has been demonstrated. The prevalence of subclinical *S. zooepidemicus* endometritis in the mare is still to be determined, and in subfertile mares the prevalence has been determined to be as high as 64%. The activation of dormant bacteria, using bacterial growth mediums like *bActivate*, may improve significantly

On the other hand, *E. coli*, a Gram-negative bacteria, was associated neither with cytologi‐ cal evidence of inflammation nor with increased pH [65, 71–72, 74–75]. The absence of correlation between *E. coli* and positive cytology may be due to its inability to attract PMNs into the uterine lumen [18]. In fact, the pathogen–host relationship, and consequently the uterine inflammatory response, of *E. coli* appears to be different from that of *β-hemolytic Streptococcus* [72]. On the other hand, it appears to associate with moderate to heavy debris on cytological smears. The interpretation of cultures positive for *E. coli* is difficult, since only pure cultures and large numbers of colonies are considered significant, and an increased sensitivity seems obtainable by culturing low volume flush [71–72, 74]. *E. coli*is sometimes considered a contaminant, but it is believed to cause subfertility in the mare [65], and it has been associated with focal infections, in the form of granulomatous plaques identified with endoscopic evaluation in two mares [72]. Another pathogen is *P. aeruginosa*, a Gramnegative bacteria, which produces a biofilm, i. e. , an adhesive matrix that harbors bacterial microcolonies and resists as well to antibiotics as to the immune system [67]. *P. aeruginosa* is

The role of non-pathogenic bacteria isolated from the mares' uterus is still unknown, but they may be the cause of decreased pregnancy rates [75]. Mares with bacteriological positivity but no cytological evidence of PMNs may be affected by contaminants [58, 72, 75]. However, their pregnancy rate resulted lower than those of the mares without any positivity to the tests. For example, *Bacillus* and *Micrococcus* are skin commensals that are not usually the cause of lowered pregnancy rates, but in barren and older mares with weakened physical barriers they may play a role in decreased fertility [75]. Even if *S. zooepidemicus* is a commensal and common

more often associated with positive cytology, differently from other pathogens [74].

the sensitivity of traditional diagnostics [76].

300 Genital Infections and Infertility

believed to be cause of chronic and persistent infections [67].



**Table 1.** Prevalence of the most common bacterial species isolated from mares' uterus in various geographical areas

#### **3. 9. Endometrial histology**

In brief, the histological grading system for the endometrium ranges from I to III. Grade I is an essentially normal endometrium with minimal alterations. Grade II is a wide category, often divided into subcategories, IIA and IIB, from mild to moderate pathologic conditions. Grade III includes severe and widespread inflammatory and/or degenerative changes of the endo‐ metrium [77]. This technique is considered the "gold standard" to diagnose endometritis, in particular the detection of PMNs infiltrating the luminal epithelium and the stratum compac‐ tum [71]. Grade IIB and III mares have also shown an increased incidence of intrauterine fluid retention, compared to grade I and IIA mares [61].

**Bacterial Species Prevalence (%) SamplingMethod Area Reference**

*Bacillus* spp. 10 Cytobrush Poland [58]

*E. coli* 10.9 Cotton Swab / Germany [65]

*E. coli* Not Knudsen Germany [74]

**Table 1.** Prevalence of the most common bacterial species isolated from mares' uterus in various geographical areas

In brief, the histological grading system for the endometrium ranges from I to III. Grade I is an essentially normal endometrium with minimal alterations. Grade II is a wide category, often divided into subcategories, IIA and IIB, from mild to moderate pathologic conditions. Grade

Biopsy

Cotton Swab / Low Volume Flush

Canada [16]

*S. equi* subsp*. zooepidemicus* 8 *E. coli* + *β-Streptococcus* 4

302 Genital Infections and Infertility

*E. coli* + *β-Streptococcus* 3

*S. equi* subsp. *zooepidemicus* 6 *E. coli* 5

*E. coli* 5

*Klebsiella* spp. Not *Pseudomonas* spp. Specified

*E. coli* Specified

Non-hemolytic *Klebsiella* spp. Not *Pseudomonas* spp. Specified

*Klebsiella* spp. Not *Pseudomonas* spp. Specified

**3. 9. Endometrial histology**

*S. equi* subsp*. zooepidemicus*

*S. equi* subsp*. zooepidemicus*

*S. equi* subsp. zooepidemicus Not Specified

*E. coli* 12 Endometrial *S. equi* subsp*. zooepidemicus* 7 Biopsy

*Bacillus* spp. 11 Endometrial *S. equi* subsp. *zooepidemicus* 8 Biopsy

*S. equi* subsp. *zooepidemicus* 9.1 Cytobrush /Endometrial

Non hemolytic Specified Catheter

*S. equi subsp. zooepidemicus* Not Cytobrush

*E. coli* Not Cotton Swab

Non-hemolytic Specified

The endometrial lesions are further classified as inflammatory, i. e. , acute, sub-acute or chronic, and non-inflammatory lesions, e. g. , hypoplasia, hyperplasia or chronic degenerative condi‐ tions. The acute inflammatory lesions are characterized by at least one PMN per 5 HPF, whereas the chronic ones by lymphocytes are often accompanied by eosinophils [58, 71]. Anyway, the number of PMNs is affected by the phase of the estrus cycle, with more neutro‐ phils especially during estrus, regardless the cause of endometritis, breeding as well as bacterial-induced. Degenerative processes are the sclerosis of uterine vessels, i. e. , "angiosis" also known as "pregnancy sclerosis" in humans, which increase with parity, the lymphan‐ giectasia, secondary to vascular degeneration or persistent inflammation, the loss of epitheli‐ um and the epithelial hyperplasia, which are highly influenced by ageing, hormones and parity [57, 61, 70]. Even if the age combined with parity has been related with endometrial degener‐ ation, and that increased parity is clearly positively correlated with increased age, old maiden mares may display much more marked changes than expected for their age [17, 61]. Sports mares differ in regard to their endometrial pathology from equine populations of nearly exclusively non-sports mares, i. e. , sports mares have a much higher prevalence of endometrial periglandular fibrosis [78]. An irregular glandular differentiation may occur physiologically during the transitional cycles, but when these findings are seen during the breeding season, they represent a pathological alteration correlated with a permanent or temporary reduced fertility [78].

In mares susceptible to post-breeding endometritis, findings may differ depending on when the examination is performed [57]. Prior to the first breeding of the year, endometrial biopsy score may be a category IIA, with pathological findings of mild, focal, subacute inflammation with or without lymphangiectasia [57]. Bacteria and PMNs are usually absent. After repeated mating in a season, these mares accumulate intrauterine fluid and may have interstitial edema within the uterine wall after ovulation. At this time, the endometrial biopsy score may worsen to a category IIB, with the primary lesions of diffuse, moderate, subacute inflammation, lymphangiectasia, and moderate to severe edema [57]. Then, positive uterine culture and neutrophilic uterine cytology can be observed [17].

#### **3. 10. Novel and old biomarkers for endometritis: myeloperoxidase and nitric oxide**

Myeloperoxidase is a pro-oxidant enzyme stored in and released by neutrophils during degranulation or after lysis [59]. In horses, myeloperoxidase has been demonstrated in different biological samples, such as plasma and broncho-alveolar lavage fluid, reaching variable concentrations. Its presence has been recently confirmed in the uterine lumen, both in physiological conditions, like during estrus, and it seems to be related to uterine inflamma‐ tion [59]. Samples were collected by low volume flush and stored in tubes with EDTA, since it prevents coagulation and consequently degranulation of neutrophils, thus stabilizing myeloperoxidase concentration. It has been noted that its concentration is high in mares affected by endometritis, showing a positive correlation both with positive cytological results and with the presence of intrauterine fluid [59]. Further studies to establish the threshold between normal and pathologic uterine concentrations of myeloperoxidase are needed.

Another intriguing biomarker is NO, a smooth muscle relaxant able to compromise the uterine contractility and, consequently, its clearance [60]. A higher amount of NO and a higher NOS expression in uterine biopsies were found in susceptible mares 13 hours after insemination, compared with resistant ones [26]. Although it is not clear whether the higher NO concentra‐ tion in susceptible mares is the cause or the result of a delayed uterine clearance, the difference between the susceptible and the resistant mares suggests a possible role for it either directly or through a NO-associated pathway [26].
