Mares presenting with intra-uterine fluid pre- and/or post-breeding can be challenging for the practitioner to deal with. It is vital to understand the underlying pathophysiology so that appropriate treatment is instigated to ensure every chance of a successful pregnancy.
Why does a mare present with intra-uterine fluid?
Mares susceptible to persistent mating-induced or infectious endometritis with delayed uterine clearance, abnormal reproductive anatomy, or mares with failure of cervical relaxation, commonly present with excessive intrauterine fluid either pre-or post-breeding.
During oestrus, it may be normal for a mare to present with a small amount (<1 cm) of anechoic intrauterine fluid; however, any fluid during dioestrus, an increased volume during oestrus or an increased echogenicity of fluid should be investigated as this may be suggestive of underlying pathology (Figure 1). An excessive oedema pattern should also be investigated, as this is likely to suggest that the mare will have an exaggerated response to semen deposition (Figure 2).
Endometritis is a leading cause of infertility in the mare, especially in older and multiparous mares (Woodward et al, 2012; Katila and Ferreira-Dias, 2022). The cause may be an acute or chronic infection or inflammatory in origin. Streptococcus zooepidemicus and Escherichia coli (Rathbone et al, 2023) are the most common aetiological agents of infectious endometritis. This infection often occurs as the result of faecal and genital flora contamination following compromise of the integrity of anatomical structures and failure to clear the uterus of bacteria because of a compromised uterine defence system, resulting in susceptibility to chronic infection.
In cases of persistent mating-induced endometritis, the inflammatory response is because of spermatozoa triggering an influx of polymorphonuclear neutrophils into the uterine lumen via activation of complement (Troedsson et al, 1998; Woodward et al, 2013). Some degree of post-breeding inflammation is to be expected, as it is a normal inflammatory response and vital for the removal of non-viable sperm and bacteria. This transient inflammation should resolve within 24 hours in mares ‘resistant’ to persistent mating-induced endometritis, so any fluid beyond 36 hours is abnormal. Mares ‘susceptible’ to persistent mating-induced endometritis have been shown to have an imbalance of pro- and antiinflammatory cytokines with increased and persistent expression of pro-inflammatory cytokines (IL-1B, IL-8, TNF-α) at 6 hours after breeding, compared to resistant mares, who have increased expression of anti-inflammatory cytokines (IL-6, IL-10, IL-IRN) (Christoffersen et al, 2012). In addition, neutrophils from susceptible mares have reduced in vitro ability to phagocytise bacteria. These susceptible mares often also have intrinsic defects in myometrial contractility, lymphatic stasis, other anatomical issues and failure of cervical dilation which further reduces their ability to clear excessive fluid (Canisso et al, 2020). Persistent inflammation, excessive accumulation of polymorphonuclear neutrophils in the uterus beyond 48 hours after breeding, intrauterine fluid and/or chronic infection will result in an increased risk of embryonic death (Canisso et al, 2020).
Diagnosing a mare with intrauterine fluid: is it infectious, inflammatory or both?
It is important to take a full history of the mare including:
A general physical examination should be performed before concentrating on the reproductive tract, paying particular attention to body condition score, any signs of pain, lameness, systemic disease or endocrinopathies. For example, mares with orthopaedic pain resulting in reduced ambulation may be more prone to fluid accumulation. Swift et al (2020) compared mares who were exercised with control mares on box rest, and found that the mares that were exercised were significantly more likely to clear intrauterine fluid.
Careful assessment of perineal conformation, vaginal speculum examination and both visual and digital examination of the cervix are vital to determine the integrity and functionality of the three anatomic barriers (vulva, vestibulovaginal fold and cervix) that prevent contamination of the uterus. Surgical solutions to correct abnormal perineal conformation include Caslicks vulvoplasty and perineal body reconstruction (vestibuloplasty, Gadd and Pouret's techniques). It is imperative to rule out urine as the source of fluid accumulation on ultrasound examinations. Mares that pool urine will have a continuous cycle of vaginitis, cervicitis and endometritis and are unlikely to conceive until the underlying cause has been resolved. Surgical solutions for urine pooling include a urethral extension and Pourets procedure. Older maiden mares may present with failure of cervical relaxation so any intrauterine fluid that forms is physically unable to be expelled. Particular attention should be paid to the cervix of these mares (and mares with a history of cervical trauma) during oestrus to determine if the cervix is relaxing appropriately.
Transrectal palpation of the reproductive tract is important before ultrasound examination to determine the location of the cervix and uterus compared to the pelvis, ie how pendulous the uterus is. Older multiparous mares generally have more gravid reproductive tracts which is likely to make them more prone to fluid accumulation.
Ultrasound examination will then help to determine the presence of any cervical, uterine or ovarian pathology. It is important to determine if any uterine oedema present is appropriate for the stage of the cycle or potentially excessive, suggestive of uterine infection or inflammation. Uterine oedema is a normal response to hormone (oestradiol) secretion when a mare is in season and will gradually increase in line with the development of the dominant follicle. Mares in oestrus will display a characteristic ‘sliced orange’ or ‘spoke wheel’ appearance on cross-sectional ultrasound examination of the uterus. Scoring systems have been developed to monitor the degree of uterine oedema, ranging from 0 (typical of mares in dioestrus) to 3 (peak uterine oedema seen during oestrus). Oestradiol secretion and maximal uterine oedema usually peak 1–2 days before ovulation, with a subsequent decrease in oestradiol reflected by a synchronous decrease in uterine oedema on ultrasound examination. Excessive or inappropriate uterine oedema patterns may indicate an underlying inflammatory or infectious process. Mares with excessive uterine oedema before breeding are at risk of excessive inflammation and subsequent intrauterine fluid accumulation post breeding. The volume, distribution and clarity of the fluid should be noted.
Tips for sampling mares with fluid to improve sensitivity and specificity
At what stage of the oestrous cycle should a sample be taken?
Taking a sample from a mare too early in the oestrous cycle may risk iatrogenic infection not being cleared from the uterus if the cervix is closed. To ensure a meaningful, reliable result, samples should be taken when the mare has at least moderate uterine oedema, a relaxed cervix and medium-sized follicles.
Aseptic preparation
Ensure the tail is held or tied out of the way and any stray hairs are contained. The vulva should be liberally scrubbed with water or very mild detergent soap. It is important when washing to begin in the centre where the labia oppose and work outwards so as not to contaminate the vaginal opening. The vulva should be dried in the same manner. Before taking the sample, it is important to ensure that there has been no inadvertent contamination of the vestibule or vagina when washing the mare or from previous rectal examination. It is also important to dry the tail and perineum so that water does not drip onto the prepared vulva.
Reducing contamination from the vaginal flora
When culturing the uterus, it is important to prevent contamination of the sample from the vagina. This can be achieved using an unguarded or guarded swab via a speculum or, ideally, with a double-guarded swab. Blanchard et al (1981) compared a partially guarded and completely guarded swab and found that the completely guarded swabs yielded significantly fewer positive cultures than the partially guarded; after a 24-hour incubation in blood agar, 25% vs 54% cultures were positive respectively. Spilker et al (2017) also strongly recommended endometrial swabbing by instrumental techniques (speculum and forceps, or iVetscope) to reduce bacterial contamination.
Double-guarded swabs are made of three parts: the sterile swab, which is surrounded by a plastic inner guard and a plastic outer guard. The outer guard is placed in the cervix before advancing the inner guard and then finally the sterile swab into the uterus. It is important to rotate the swab for a minimum of 30 seconds to ensure a representative sample is obtained. The process is reversed to prevent contamination on the way out. If a separate cytology brush is being used, the sterile cytology brush and inner guard can be inserted through the swab's outer guard (after removing the sterile swab and inner guard) to prevent further contamination from the vaginal flora. In mares that are pooling fluid, it may be beneficial to guide the swab ventrally via the rectum or to aspirate a sample of fluid directly via an artificial insemination pipette.
Sample submission
If the swab cannot be plated immediately, it should be carefully placed in an appropriate transport medium, refrigerated and transferred within a couple of hours. Overgrowth of bacteria in the transfer media may result in a false positive result. Taylorella equigenitalis is very fragile and sensitive to temperature and light, so Amies transport medium with charcoal is required to prevent false negative results.
It is always best to submit the swab immediately after sampling. In general, no more than 24 hours (48 hours at the absolute maximum) should elapse before starting bacteriological examination and no more than 48 hours should elapse before polymerase chain reaction testing.
Endometrial flush cultures
The use of a ‘low-volume’ uterine lavage may be indicated to increase the sensitivity and specificity of obtaining an endometrial sample for culture. A low-volume lavage is considered to be more sensitive as fluid spreads on a large endometrial surface and therefore is thought to be more representative of the entire uterus. Cocchia et al (2012) diagnosed acute endometritis in 50% (10/20) of mares sampled using a double-guarded cotton swab, 25% (5/20) sampled with a cytobrush and 75% (15/20) sampled by low-volume flush. Le Blanc et al (2007) found the isolation of micro-organisms to be highly associated with cloudy and mucoid effluxes, increased efflux pH, debris on cytology and polymorphonuclear neutrophils on biopsy (Figure 3). Therefore, the appearance of the efflux is important in differentiating contaminants from meaningful growth.
A sterile lavage tube connected to a 500 ml or 1 litre bag of sterile saline is used as a closed system to infuse and recover fluid from the uterus. The fluid bag is then hung up to settle for a couple of hours before aspirating two 50 ml samples into two Falcon tubes. The Falcon tubes are then centrifuged to obtain two pellets of cells, one for culture and one for cytology. When this method was first described, only 50 ml of saline was typically infused (hence a ‘low-volume’ lavage); however, such a low volume can be very difficult to retrieve and, in older multiparous mares, may not distribute throughout the entire uterus so is less likely to provide a representative sample. As such, the author prefers to use either a 500 ml or 1 litre saline bag.
Von Dollen et al (2021) used 3.3% of N-acetyl-cysteine infusion to increase the yield of positive cytology and culture. Following infusion of 6 g of N-acetyl-cysteine in 150 ml saline into the uterus and culture of the lavage fluid 24 hours later, an increase in positive cultures from mares which had previously been negative was seen. It is likely that certain mares produce excess mucous that protects bacteria and prevents their detection, meaning an accurate diagnosis is only possible following disruption of this mucous.
Endometrial biopsies
An endometrial biopsy can help diagnose the presence of both acute neutrophilic infiltration of the uterus and chronic lymphocytic inflammation. The presence of polymorphonuclear neutrophils on endometrial histology is often reported as the reference standard test for diagnosing equine endometritis (Nielsen, 2005). Buczkowska et al (2014) compared bacteriological and cytological results obtained from endometrial biopsies and endometrial cytology samples obtained with a cytobrush and assessed their diagnostic accuracy relative to the presence of polymorphonuclear neutrophils in endometrial tissue. The sensitivity for bacterial growth (monoculture) and cytology (>2% polymorphonuclear neutrophils) following an endometrial biopsy were 63% and 73% respectively, with a specificity of 54% and 96% (Buczkowska et al, 2014). The sensitivity of bacterial growth (monoculture) and cytology (>2% polymorphonuclear neutrophils) from a cytobrush sample were 50% and 71% respectively, with a specificity of 73% and 85% (Buczkowska et al, 2014). Overbeck et al (2011) also compared the accuracy of a uterine swab, cytological brush and endometrial biopsy for the detection of subclinical endometritis in the mare, using the presence of polymorphonuclear neutrophils on endometrial biopsy as the reference standard test. At a threshold of >2% polymorphonuclear neutrophils, the sensitivity for cytology was 17% for cytological brush, 0% for swab and 25% for biopsy. In the same study, the specificity for cytology was 83% for cytobrush, 93% for swab and 85% for biopsy. The reported sensitivity for culture of a cytological brush was 25%, 33% for swab and 25% for biopsy (Overbeck et al, 2011).
Marth et al (2018) evaluated the expression of innate immune genes in endometrial biopsies and found the expression of genes for the antimicrobial peptides equine β-defensin 1, lysozyme and secretory leukoprotease inhibitor was higher in mares susceptible to persistent mating-induced endometritis than control mares. The diagnostic sensitivity of assays for equine β-defensin 1, lysozyme and secretory leukoprotease inhibitor gene expression to detect susceptibility to persistant mating-induced endometritis was estimated to be 100%, 94% and 100% respectively, with specificities of 83%, 78% and 78% . When all three tests were positive, the specificity increased to 94%, with an overall sensitivity of 94%. While these tests are not currently available commercially, they may help in the differentiation of resistant and susceptible mares which will further aid future therapeutic choices.
Endometrial biopsy samples should be either fixed in Bouin's for 24 hours or formal saline and then transferred to formal saline. Again, if sent via the post, ensure the sample will arrive at the lab within 24 hours if placing in Bouin's, otherwise place directly in formalin.
Treatment
Appropriate treatment should aim to improve physical clearance of inflammatory by-products following breeding, eliminate any potential pathogens (mechanical pathway), modulate the ongoing inflammatory response (immunological pathway) and correct any anatomical abnormalities.
Elimination of potential pathogens
In a time of increasing antimicrobial resistance, the judicious use of antimicrobials in equine reproductive practice is crucial. Empirically choosing antimicrobials for the treatment of endometritis and the routine use of post-breeding infusions should be avoided. Reporting changes in antimicrobial resistance over time can help clinicians make informed decisions and help conserve critical antimicrobials for future use.
Rathbone et al (2023) retrospectively described temporal changes in antimicrobial resistance of endometrial bacteria isolated from Thoroughbred broodmares in England between 2014 and 2020. They found beta-haemolytic Streptococcus (52.5%) and E. coli (25.8%) to be the most frequently isolated pathogens, and that there was no increase in antimicrobial resistance in beta-haemolytic Streptococcus or E. coli to penicillin (99.6% beta-haemolytic Streptococcus-susceptible), gentamycin (81.7% E. coli-susceptible) and ceftiofur.
Mitchell et al (2018) reviewed bacterial isolates submitted to Cornell University between 2007–2017 and their respective antimicrobial sensitivities to determine if resistance patterns had changed over time. Of 2311 samples submitted, 912 (39.4%) were culture positive. Of the positive samples, 30% were as a result of infection with E. coli, 29% beta-haemolytic Streptococcus, 10% other Streptococcus spp. and 5% grew Enterococcus faecalis. The antimicrobial with the most susceptibility (>90% sensitive) among the top three isolates was trimethoprim-potentiated sulphonamides and the second most susceptible (>80% sensitive) antimicrobial was amikacin.
While suitable intra-uterine and/or systemic antimicrobials following sensitivity testing of uterine pathogens are the most effective treatment against infectious endometritis, antimicrobials alone may be insufficient to eliminate bacteria from the uterus. Suggested presence of biofilms, dormant bacteria and mucous may prevent antimicrobials from reaching their target and effectively eliminating bacteria.
Approximately 80% of bacteria isolated from the equine uterus is capable of producing a biofilm (Ferris, 2014; Ferris et al, 2016). Host immunity and the microenvironment are thought to play a role in biofilm formation, but the implication of a biofilm on the pathogenesis of endometritis in the equine uterus is yet to be fully understood.
Uterine lavage with various mucolytics such as a 1% hydrogen peroxide, 20% N-acetyl-cysteine or 99% dimethyl sulfoxide solution may be indicated (Table 1). BActivate has been suggested to activate subclinical bacterial endometritis caused by beta-haemolytic Streptococcus, allowing for effective antimicrobial treatment, resulting in increased pregnancy rate/cycle and lower fetal loss in mares that have issues maintaining a pregnancy (Petersen and Bojesen, 2019).
| Compound | Dose | Administration | Indications/notes |
|---|---|---|---|
| Dimethyl sulfoxide 99% | 50–200 ml | Diluted in 1 litre sterile saline for infusion followed by lavage with saline or Hartmann's in 24 hours | Anti-inflammatory and oxygen scavenger Mucolytic |
| Hydrogen peroxide 6% | 60–120 ml or 1% solution as lavage fluid | Infusion, followed by sterile saline or Hartmann's lavage in 24 hours | Treatment of fungal endometritis Biofilm disruption* |
| N-acetyl-cysteine 20% solution | 6 g (30 ml) of 20% solution | 30 ml diluted in 150 ml sterile saline (3.3%) for infusion, followed by sterile saline or Hartmann's lavage in 24 hours | Mucolytic |
| Povidone-Iodine | 0.01% | 10 ml added to 1 litre sterile saline | Bacterial and fungal endometritis |
| Tris-EDTA | 250–500 ml | Infusion followed by lavage with Lactated Ringers Solution | Can be used in combination with antimicrobials (enhances bacterial killing) |
In addition, correction of perineal conformation (such as Caslick's vulvoplasty, vestibuloplasty, Gadd or Pouret's procedure) and cervical repair, if indicated, is imperative to prevent continued contamination and pneumovagina.
Modulating the inflammatory response and improving mechanical clearance
Uterine lavage with pH balanced, warm fluids such as Compound Sodium Lactate (Hartmann's solution) or Lactated Ringers Solution will physically reduce the inflammatory exudate and bacteria numbers, stimulate uterine contractions by stretching the uterus and recruit viable neutrophil influx by transient irritation of the uterus. Ideally, in mares known to be susceptible to persistant mating-induced endometritis, lavage should be performed 4–6 hours post-breeding as this is associated with increased pregnancy rates compared to waiting until 12–24 hours post-breeding (Brinsko et al, 1991).
Uterine lavage in combination with ecbolics, such as oxytocin and cloprostenol, remain the mainstay of treatment. Oxytocin (10–20 IU intramuscularly or intravenously from 4 hours post breeding) provides high amplitude uterine contractions for a limited duration and requires repeated doses because of its short half-life (Brinsko et al, 1991). Oxytocin has also been shown to improve pregnancy rates in sub-fertile mares (Scarlet et al, 2023). Treatment is typically discontinued after 72 hours post ovulation because of cervical closure, but minimal amounts of fluid will still be removed after this via the uterine lymphatics.
Cloprostenol (Estrumate, 250 mcg intramuscularly every 24 hours) provides lower amplitude, more sustained contractions but is only advisable pre-ovulation; if given post ovulation, luteal function may be impaired for several days (Nie et al, 2002). In addition, correction of perineal conformation (such as Caslick's vulvoplasty or Gad's procedure) and cervical repair, if indicated, is imperative to prevent continued contamination and pneumovagina.
Swift et al (2020) reviewed the effect of exercise, oxytocin and cloprostenol on fluid accumulation and found that compared to the box rest negative control mares, exercise was the most effective treatment with 29.7 times increased odds of fluid clearance. The second most effective treatment was oxytocin alone, followed by cloprostenol. The combination of exercise and oxytocin had 8.4 times greater odds of fluid clearance compared to control mares. This study highlights the importance of exercise where at all possible in these mares.
Käser et al (2016) showed a significant increase in endogenous oxytocin levels in response to teasing and, if possible, exposure to a stallion while the mare is still responsive can aid fluid clearance further. Various immunomodulatory treatments for persistent mating-induced endometritis have been studied. Bucca et al (2008) found that a single dose of 50 mg intravenous dexamethasone at breeding significantly improved per cycle pregnancy rates in mares accumulating ≥2 cm of endometrial fluid post-breeding and in mares where fluid persisted 36 hours after mating. Dexamethasone was shown to reduce pro-inflammatory cytokines, suppression inflammatory mediators and increase anti-inflam-matory cytokines. However, it may not always be appropriate to give a mare a high dose of steroids because of the potential risk of inducing laminitis, especially in horses with a previous history of laminitis (Johnson et al, 2002).
Both phenylbutazone and flunixin meglumine non-steroidal anti-inflammatory drugs have the potential to delay uterine clearance because of prostaglandin inhibition and can have negative effects on ovulation (Cuervo-Arango, 2011; Lima et al 2015). Firocoxib, a specific cyclo-oxygenase-2 inhibitor, has been shown to reduce endometrial inflammation in mares susceptible to persistent mating-induced endometritis with no adverse effects on ovulation rate or embryo recovery when 0.2 mg/kg of firocoxib was given daily from ovulation induction to the day after insemination (Friso et al, 2019).
Segabinazzi et al (2017) studied the effects of platelet-rich plasma on uterine inflammation, conception rates, endometrial polymorphonuclear leukocyte migration and cyclo-oxygenase-2 protein levels in endometrial tissue. Platelet-rich plasma treatments resulted in a decrease of polymorphonuclear leukocytes in the cytology after breeding when compared to controls. Intrauterine fluid accumulation did not differ between cycles; however, the conception rates were significantly higher in the mares treated with platelet-rich plasma. This study concluded that platelet-rich plasma beneficially reduces inflammatory response in mares with persistent mating-induced endometritis independent of when treatments were administered, thus increasing the chances of a successful pregnancy. Early work has also suggested that pentoxifylline may improve post-breeding uterine clearance and embryo recovery rate (Oliveria et al, 2022) but further controlled studies are required.
The ability of mycobacterial cell wall extract to stimulate the induction of cytokines and chemokines, activate neutrophils and enhance the overall host immunological response is well documented (Fumuso et al, 2007; Woodward et al, 2015). Rogan et al (2007) reported that the use of 1.5 ml by intrauterine or intravenous administration during early oestrus in mares experimentally induced with beta-haemolytic Streptococcus endometritis enhanced the innate humoral immune response. The overall efficacy of mycobacterial cell wall extract to eliminate beta-haemolytic Streptococcus infection compared with placebo in this study was 35% by day 1 post-ovulation and 70% by day 7 post-ovulation.
Conclusions
An accurate diagnosis and understanding of the underlying pathophysiology of the fluid-producing mare is vital to ensure appropriate and timely treatment is instigated. Any infectious endometritis must be resolved before breeding. Careful breeding management should not be underestimated to ensure appropriate timing of breeding in relation to ovulation. Uterine lavage and administration of ecbolic agents 4–6 hours post insemination in mares susceptible to intrauterine fluid accumulation should be performed when possible. Targeted post-breeding treatments to modulate the mare's ongoing immune response are critical. The benefit of both exercise and teasing where possible should not be underestimated, and any anatomical issues must be corrected. In order to provide the best chance of a successful pregnancy, it is vital to have a non-infected and un-inflamed uterine environment.