The Use of Antibiotics in Broodmares With Post-service Endometritis A Knowledge Summary

There is currently a lack of evidence to suggest that antibiotics delivered into the uterus post-service improves pregnancy rates. The evidence does support that post-service treatment with or without antibiotics improves pregnancy rates in mares with endometritis compared to no treatment (pregnancy rates between 40-90% compared to 0-56% respectively). The studies further identify the need to recognise mares at risk of post-service endometritis to target appropriate therapy.


The Evidence
Seven publications were identified in the literature searches.These comprised 2 descriptive studies, 1 observational study, 2 randomised non-blinded controlled trials and 2 non-randomised non-blinded controlled trials.Included publications were of level 3 (randomised control trials) or level 4 (lower quality controlled trial) (OCEBM Levels of Evidence Working Group).The outcomes studied varied between the studies including pregnancy rates at 15 days (Pycock and Newcombe 1996;Gores-Lindholm et.al. 2013;Sharma et.al. 2011), 45 days (El-Roos 2004), 60 days (Noseir 2002;Taha 2007), or unknown (Pascoe 1990).Further, there was no standard population treated, antibiotic prescribed or method of administration between the studies.In the three studies with no case selection any treatment was seen to benefit pregnancy rates over no treatment at all (Pycock and Newcombe 1996; Pascoe 1990; El-Roos 2004).This is similar to mares with a previous history of endometritis with a significantly higher pregnancy rate in mares treated with antibiotics and ecbolic agents compared to no treatment at all (Noseir 2002;Taha 2007).In the two studies which compared mares with and without signs of endometritis and treatment with antibiotics, higher pregnancy rates were achieved in mares that were not treated and had no signs of endometritis (Sharma et al., 2011;Gores-Lindholm et.al., 2013).The findings support the need for case selection in treating mares with postservice induced endometritis.

Clinical bottom line
There is currently a lack of evidence to suggest that antibiotics delivered into the uterus post-service improves pregnancy rates.The evidence does support that post-service treatment with or without antibiotics improves pregnancy rates in mares with endometritis compared to no treatment (pregnancy rates between 40-90% compared to 0-56% respectively).The studies further identify the need to recognise mares at risk of post-service endometritis to target appropriate therapy.
Intervention details: Mares were assigned in rotation to one of four treatment groups.
Group 1: no treatment (n=322) Group 2: intrauterine antibiotic treatment within 72 hours of mating (1200mg procaine penicillin combined with 1500mg dihydrostreptomycin and 900mgs framycetin made to a volume of 12 mls or neomycin (1g) polymyxin B (40,000 iu) and furaltadone (600mg) with crystalline benzylpenicillin (3g).(n=317) Group 3: oxytocin (25 iu) administered by IV injection within 72 hours of mating (n=316) Group 4: oxytocin (25 iu) administered IV followed 30mins later by intrauterine antibiotics as described for group 2 within 72 hours after mating (n=314) Study design: Prospective non-blinded, non-randomised controlled-trial Outcome studied: Ultrasound pregnancy detection on day 13, 14, or 15 post ovulation and repeat examination between 27-30 days after ovulation.Early embryonic death rate determined by the presence of a vesicle at first scan and absence at second scan

Main findings: (relevant to PICO question):
-Any treatment (oxytocin alone or in combination with intrauterine antibiotics) significantly increased pregnancy rates over untreated control mare (control pregnancy rate 56%) -No difference in pregnancy rates between group 2 (intrauterine antibiotics) (64%) and group 3 (oxytocin) mares (63%) -Mares in group 4 (intrauterine antibiotics and oxytocin) had higher pregnancy rates than mares in group 2 and group 3 (group 4 pregnancy rate 72%) -No difference in early embryonic loss between treatment groups Limitations: -Mares were not subject to any diagnostic procedure to confirm the presence or absence of infection prior to antibiotic treatment -There was no data pertaining to presence of clinical or subclinical signs of endometritis such as intraluminal fluid within the groups prior to treatment -Unknown spread of ages and reproductive status (barren, maiden, post-parturient) within groups -Unknown if mares were selected at foal heat or recruited to the study after multiple mating attempts during the season

Main findings: (relevant to PICO question):
-There was a significantly higher pregnancy rate in mares that had a uterine lavage containing antibiotics and an injection of prostaglandin (83.3%) compared to other treatment groups -There was no difference in pregnancy rates between mares receiving antibiotic lavage alone (50%) or combined with oxytocin (66.6%) -All treatments significantly increased pregnancy rates over control (untreated) mares (0%) Limitations: -No comparison was made to treatment without antibiotics such as oxytocin alone -No description of how randomisation was performed and if mares were randomised based on criteria for selection into the study between treatment groups -Unusual antibiotic delivery through lavage system and unusual cytology performed from antibiotic flush -No cytology or culture performed prior to antibiotic lavage -No evidence of other complicating factors such as presence of fluid or uterine oedema

Sample size: n = 154
Intervention details: Mares were client owned and divided into groups based on endometrial swab and culture results obtained pre-breeding.Group 1 (n=44): repeat breeders that had clinical findings suggestive of mucus over-secretion with no evidence of bacterial infection from culture of uterine lavage fluids.Mares were treated with Nacetylcysteine (NAC) 24-36 hours prior to cover followed by oxytocin (20 iu IM) 4-8 hours after infusion Group 2 (n=85): control group of reproductively normal mares presented on their first cycle showing no evidence of infection or inflammation on swab and cytology and intra-uterine fluid <5mm Group 3 (n=25): mares that had previous positive uterine cultures treated with lavage and antibiotics on previous cycle and no evidence of mucus over-secretion, intrauterine fluid >1cm Mares within these groups were additionally treated with intrauterine antibiotics (1g ceftiofur) dependent on previous history and farm management preferences.

Study design: Descriptive study
Outcome studied: First cycle pregnancy rate between 13 and 15 days

Main findings: (relevant to PICO question):
-In mares that were suggestive to have mucous over secretion (group 1) there was no significant difference in first cycle pregnancy rates between mares receiving antibiotics and NAC and those not receiving antibiotics (80% and 71% respectively) -Similarly, in mares that had previously been treated for endometritis (group 3) there was no significant difference in pregnancy rates between mares treated with lavage and oxytocin, and lavage, oxytocin and antibiotics (50% and 62% respectively) -In mares with no history of endometritis (group 2) the pregnancy rate was higher for mares that were lavaged with LRS and given oxytocin (89%) than those additionally treated with antibiotics (60%) Limitations: -The study was aimed at assessing the potential for NAC to be used to clear mucous and was not comparing antibiotics as post-service treatment -The study was not blinded and it is not possible to tell why antibiotics were used or not used -Mares that previously swabbed clean may have been better identified with a low-volume lavage than uterine swab -No separation of mare age or status -In group 2 mares that had no evidence of infection, 80% of those treated with antibiotics had greater than 1cm of fluid or marked oedema 18-24 hours post-mating.There was only 18% of mares with greater than >1cm of fluid in mares treated with LRS and oxytocin alone

Noseir (2002)
Population: Thoroughbred mares aged 10-15 years with a history of recurrent and persistent endometritis

Main findings: (relevant to PICO question):
There was no significant difference in pregnancy rates between mares receiving intrauterine antibiotics alone (57.3%)or oxytocin administration alone (52.5%).There was, however, a significantly higher pregnancy rate between mares receiving antibiotics combined with oxytocin (67.8%) compared to other treatment groups.There was a significantly higher pregnancy rate in all treatment groups compared to the control (22.5%).

Main findings: (relevant to PICO question):
Higher pregnancy rates were achieved with the combination of leukocyte enriched plasma and oxytocin (72%), compared to leukocyte enriched plasma and antibiotics (61%) Limitations: -Unknown randomisation method, low statistical power, and lack of statistical tests performed -Mares received double guarded swabs for culture prior to cover; however a mare with a positive culture (current endometritis) was not excluded from the trial.The distribution of mares with endometritis prior to cover was unknown between the groups -Unknown presence of clinical or subclinical signs of endometritis between groups -Unknown age distribution of mares and inclusion of maiden, barren and foal heat mares

Main findings: (relevant to PICO question):
Significantly higher (p<0.05)day 39 pregnancy rates and foaling rates observed in untreated oestrous cycles (46.39%) than treated cycles (40.62%).No difference in pregnancy rates between treated oestrous cycle and untreated oestrous cycle at day 16 of pregnancy (46% and 51% respectively) The incidence of uterine therapies increased with increasing age Limitations: -There were no criteria defined to determine which treatment option was chosen and the duration of treatment -Antibiotic choice and method of administration not discussed -Unclear at what time antibiotics were used post ovulation and if other methods were used prior to cover -No prior history of mares to determine spread of barren, maiden or foaling mares and if endometritis was present prior to cover

Appraisal, application and reflection
Post-service endometritis is a common condition and can be challenging for the clinician to diagnose and treat.Current treatment options for endometritis include intrauterine antibiotics, lavages (LRS or antibiotic infused), other uterine infusions such as mucolytics (NAC, EDTA), immune modulation treatments (plasma), and ecbolic agents (oxytocin and / or prostaglandin).The purpose of this Knowledge Summary was to evaluate the evidence comparing the use of intrauterine antibiotics in the treatment of endometritis to other routine therapies.
The diagnosis of post-service endometritis can be challenging for the clinician and varied between the studies.4/7 studies compared treatments in mares affected by endometritis whilst the remaining studies Intrauterine antibiotics have been postulated to not only aid in the treatment of potential pathogens but also by cervical dilation enhancing fluid removal (Pycock and Newcombe 1996).The combination of oxytocin to stimulate uterine evacuation and manual dilation of the cervix has been similarly treated with the use of plasma and uterine lavages (Sharma et.al. 2011; Gores-Lindholm 2013; El-Roos 2004).Gores-Lindholm (2013) demonstrated no significant difference in pregnancy rates for mares previously treated for endometritis with lavage and oxytocin (50%) compared to lavage, oxytocin and antibiotics (62%).Similarly, Sharma et.al (2011) showed no difference in day 16 pregnancy rates in mares treated with uterine lavage and oxytocin (52%) and intra-uterine antibiotics and oxytocin (45%).While the type, volume of antibiotic and method of antibiotic installation varied between the studies, all involved uterine placement which would have led to cervical dilation.
Supporting the use of antibiotics are culture and sensitivity studies.Common bacterial isolates in endometritis include beta-haemolytic streptococcus, corynebacterium spp, and Escherichia coli (Assad et.al 2015).In all studies, the choice of antibiotic was empirical and not determined by culture and sensitivity results.Four of the studies used a combination of procaine penicillin with streptomycin, gentamicin, framomycin, and or neomycin (Pycock and Newcombe 1996;Pascoe 1990;Taha 2007;Noseir 2002) whilst one study used ceftiofur alone (Gores-Lindholm 2013).Two studies did not comment on the antibiotic used (Sharma et.al., 2011 ;El-Roos 2004), and no studies commented on the use of systemic antibiotics.Whilst the types of bacteria isolated are likely to vary geographically, Assad et.al., (2015) showed all isolates were sensitive to gentamicin and ceftriazone (100%) followed by tetracycline and chloramphenicol (92.8%).There was widespread resistance against streptomycin (92.8%) and penicillin (50%) which were commonly used in the studies.Sensitivity patterns varied based on which bacteria were isolated, with streptococcus spp having maximum sensitivity to gentamycin and tetracycline antibiotics (100%) and moderate sensitivity to penicillin (71.4%), while E. coli isolates were 100% resistant to ampicillin and streptomycin.Indiscriminate use of antibiotics has shown to increase the risk of fungal and resistant bacterial species (Assad et.al. 2015).Further use of diagnostics to identify the bacterial isolate(s) involved, and antibiotic sensitivity, may aid in the reduction of restricted antibiotics such as tetracyclines and chloramphenicol.
The outcome achieved varied between the studies, with different days of gestation considered.Majority of ) commented on embryonic loss with uterine treatments by comparing pregnancy rates at 12-16 days to 28 days.There was no significant difference in embryonic loss in either study, however Pycock and Newcombe (1996) noted a higher embryonic loss when antibiotics were installed via a syringe compared to a catheter system.Early embryonic loss is estimated to be between 2.6% to 24% due to progesterone deficiency or failure of maternal recognition (Vanderwall 2008).Such that the later scanning dates in the studies by El-Roos (2004) and Taya (2007) may have had a higher negative pregnancy rate from early embryonic loss rather than failure to conceive.
Post-service endometritis is a multifactorial disease and can be challenging to manage.Currently the collective data supports the combined use of intrauterine treatments and ecbolic agents to increase pregnancy rates in mares with a history and or signs of endometritis.It is unclear if similar pregnancy results may be obtained with cervical dilation methods that do not include antibiotics in combination with ecbolic agents alone.Further, this summary identifies the need for ongoing research into defining which mares have endometritis in order to determine those that may benefit from treatment.Additional diagnostics such as culture and cytology of endometrial samples post-service could be utilised by clinicians to justify antibiotic usage.Currently clinicians should use intrauterine antibiotics judiciously as they are unlikely to improve pregnancy rates and may be detrimental to fertility in mares with no history or signs of endometritis.Inclusion: Any reported use of antibiotics used in post-service treatment from natural coverings.

p a g e | 5 total pages: 14 Study design: Prospective, randomised, non-blinded controlled trial Outcome studied:
Pregnancy rate after 2 months of treatment Taha (2007)2)ave descriptions of how they diagnosed endometritis.Sharma  et.al., (2011)andNoseir (2002)treated mares with endometritis with no definition stated.Taha (2007)defined persistent mating induced endometritis as a history of repeat breeding of more than 3 natural services without a pregnancy, intrauterine fluid one to two days after breeding, or endometrial cytology showing >5% PMN's.Gores-Lindholm et.al., (2013) diagnosed endometritis based on previous cycle uterine lavage fluid with a positive culture, repeat breeding of at least 2 cycles or 2 unsuccessful embryo transfer attempts.The lack of a consistent definition of endometritis makes comparison of treatment options challenging.The use of controls that did not receive antibiotics was highly variable amongst the studies.The most consistent comparisons used between the studies were either no treatment or ecbolic agents alone.6/7studiescompared no treatment as a control(Pycock and Newcombe 1996; Pascoe 1990; Taha 2007; Gores-Lindholm et.al.2013; Noseir 2002; Sharma et.al 2011), while 4/7 studies compared oxytocin at intervals ranging from 1-72 hours post mating [(Pycock and Newcombe 1996) 72 hours post mating, (Noseir 2002) 4-12 hours post mating, (El-Roos 2004) 1, 12, 16 hours post mating, (Sharma et.al.2011) unspecified time/s].Taya (2007) was the only study to evaluate the use of prostaglandin as an ecbolic agent.Consistently it was observed that treatment of mares with endometritis improved pregnancy rates over no treatment at all (Pycock and Newcombe 1996; Pascoe 1990; Taha 2007; Gores-Lindholm et.al.2013; Noseir 2002; Sharma et.al 2011).Ecbolic agents and antibiotics had similar improvement to pregnancy rates when used alone (Pycocok and Newcombe 1996; Noseir 2002).The combined effect of antibiotics and ecbolic agents were seen to act in synergy and further improve pregnancy rates (Pycock and Newcombe 1996; Taya 2007; Noseir 2002).

11 total pages: 14 Methodology Section
or antibiotics or antimicrobial or antimicrobials or anti-microbial or anti-microbials or antibacterial or antibacterials or 'antiinfective agent' or 'antiinfective agents' or 'anti-infective agent' or 'anti-infective agents' or penicillin or gentamicin).mp.or exp antibacterial agents/ or exp antibiotics/ or exp antiinfective agents/ 4 (Post-service or 'post service' or postservice or mating or induced or breeding).mp. 5 (pregnant or pregnancy or conception or conceiving or fertility or fertile or foal or foals).mp.or pregnancy/ or fertility/ 6 1 and 2 and 3 and 4 and 5 PubMed 1. equine OR horse OR mare OR mares OR broodmares OR brood mares OR pony OR ponies 2. endometritis or endometriosis or endometrial or metritis or uterine or uterus 3. antibiotic or antibiotics or antimicrobial or antimicrobials or anti-microbial or anti-microbials or antibacterial or antibacterials or 'antiinfective agent' or 'antiinfective agents' or 'anti-infective agent' or 'anti-infective agents' or penicillin or gentamicin 4. Post-service or post service or postservice or mating or induced or breeding 5. pregnant or pregnancy or conception or conceiving or fertility or fertile or foal or foals 6. 1 and 2 and 3 and 4 and 5 Dates searches performed: March 2017 DOI: http://dx.doi.org/10.18849/ve.v3i1.132next review date: 20 Mar 2019 Veterinary Evidence ISSN:2396-9776 Vol 3, Issue 1 DOI: http://dx.doi.org/10.18849/ve.v3i1.132next review date: 20 Mar 2019 p a g e |

p a g e | 12 total pages: 14 Exclusion / Inclusion Criteria Exclusion
: Non-English language, narrative or non-systematic review articles, unpublished data, pharmacokinetic, in vitro or in vivo experimental studies, artificial insemination protocols.