BREEDING THE PROBLEM MARE INTRODUCTION
BREEDING THE PROBLEM MARE
Management of the "problem mare" is one of the most challenging and potentially frustrating aspects of equine reproduction. There are no magical treatments despite anecdotal evidence of clients who each "once had a problem mare that remained barren until they either: pasture mated her, infused her uterus with kerosene, supplemented her with vitamin E, took 10 L of blood from her, transported her after breeding, bred her to a pony teaser stallion, or fed her extracts from the local homoeopath." The simple truth our clients need to understand is, that apart from semen, it is hard work and persistent adherence to strict scientific principles and management techniques that result in conception and maintenance of pregnancy with problem mares. This presentation will strive to amalgamate these principles into a practical approach. People's attitudes and time constraints are major difficulties of successfully breeding the problem mare. Frequently, mares are on different breeding farms and despite initial enthusiasm from all concerned, with each unsuccessful cycle a negative attitude engendered from the farm manager, veterinarian and attendants becomes apparent. As the treatments change and the mare keeps returning to oestrus, hidden pressures on the veterinarian to get to the next farm on time begin to influence judgement. Many veterinarians would gladly refer these cases to specialist facilities at the end of the season, but few think it necessary at the beginning. This chapter outlines the philosophy and treatment of the "problem mare program" at the Goulburn Valley Equine Hospital (GVEH). However the approach, medical and surgical therapies and management techniques should apply to any breeding establishment. This program has an annual enrolment of between 40 and 100 problem mares, boarded full time on pasture at the hospital. The number of mares varies according to time of year, economic trends and removal of pregnant mares and/or mares with a hopeless prognosis. Facilities are designed to minimise handling time and maximise individual mare attention. An added benefit is that, because most mares are bred at the hospital or transported to and immediately back from local farms for breeding, they can be intensively managed and the psychological oppression of mares returning to heat is generally only a local (hospital) phenomenon. The cost to the client is often greater because mares are able to be managed more intensively than on breeding farms. However, in our experience, few people complain about billing once their mare is pregnant. Although a headache, constant owner communication and education reduces the number of dissatisfied clients when mares fail to become pregnant. This represents one approach to breeding problem mares and is the combined opinions and experiences of the authors only. There are many other philosophies and personal preferences that may not be addressed. Perhaps we should keep in mind that for any condition with so many different avenues of treatment (much of it anecdotal, undocumented or not scientifically validated) that without an experimental model it is not only difficult, but inappropriate, to be dogmatic in recommendation for management.
WHAT IS A PROBLEM MARE?
"When" a mare begins to be regarded as a problem breeder is at the discretion of the owner, veterinarian or farm manager. Classically, a mare should be regarded as a potential problem if she fails to conceive to a fertile stallion on a well managed breeding farm on three or more cycles in one season. We have seen instances of reproductively normal mares failing to conceive to a subfertile stallion for as long as three breeding seasons. Unfortunately, fertility of a stallion may not be common knowledge away from the breeding farm, particularly to the mare owner. In addition, improper handling and/or insemination techniques in an artificial insemination (AI) program are excellent methods for dramatically reducing pregnancy rates. Attention to pregnancy rate per cycle from maiden, dry and wet mares may help determine if stallion fertility or management procedures should be questioned. In evaluating a problem mare, all aspects of management should be examined. This might include nutrition, teasing techniques, parasite control, medical history, evaluation for chronic generalised disease and other management practices. There are many reasons mares may become problems. For example, mares failing to exhibit oestrus in the breeding season may have a granulosa theca cell tumour, gonadal dysgenesis, pituitary adenoma, or they may be pseudopregnant, unknowingly pregnant, or in anoestrus. However, by far the most common type of problem mare presented are mares with persistent, unidentified or inappropriately treated uterine infections. All mares presented to the GVEH have been barren two or more breeding seasons and most have received some treatment in the breeding seasons prior to referral. The following discussion relates only to mares with acute and/or chronic uterine inflammatory changes.
HOW THE PROBLEM OCCURS
Recognition of factors responsible for mares susceptibility to uterine infection and persistent uterine inflammation or its sequelae should dramatically reduce the number of problem mares. However, the number of mares foaling each year has not appreciably improved in those breeds whose registries have kept accurate records for many years. This can be related partly to economic constraints and lack of genetic selection for fertility, however extremely high (> 90%) pregnancy rates with only slightly lowered foaling rates consistently occur on some well managed breeding farms. This fertility can be attributed partially to prebreeding selection of the most fertile mares and correcting faulty management procedures and non-infectious causes of infertility. However, careful management to ensure prompt treatment or prevention of uterine inflammation is most important. Mares do not become problem breeders overnight. Susceptibility to infection is a graded condition and occurs primarily due to effects of increasing age, reproductive tract damage and bacterial challenge. Bacterial challenge is influenced by external conformation, breeding techniques, examination procedures, anatomical abnormalities, and post-partum events. The physical barriers to infection are the external vaginal lips, the vestibular sphincter and the cervix. The ability to isolate bacteria in normal mares decreases progressively from the clitoral fossa (94%) to the vestibule (69%), cranial vagina (42%), and uterus (31%) (Hinrichs, Cummings, Sertich, & Kenney. 1989). In this study, no potential pathogens were isolated from vaginal or uterine cultures (Hinrichs, Cummings, Sertich, & Kenney. 1988). Contamination of the uterus with bacteria is inevitable. Potentially pathogenic organisms are introduced at breeding, during and after parturition, during examination and as a result of failure of physical barriers to infection (ie. pneumovagina). When uterine defence mechanisms are functioning properly, massive challenges, either natural or experimental, fail to produce inflammation that lasts long enough to interfere with reproduction (Asbury, Schultz, Klesius, Foster, & Washburn. 1982; Asbury, Gorman, & Foster. 1884; Evans, Hamer, Gason, Graham, Asbury, & Irvine. 1986). For embryo survival, mares must clear bacteria and inflammatory products from the uterus, as a result of breeding, by the time the embryo descends into the lumen about 5 to 6 days after ovulation (Oguri & Tsutsumi. 1972). Uterine defence mechanisms are mechanical and cellular. Mechanical contributions to uterine defence are myometrial contractions which assist in evacuation of uterine contents and a relaxed cervix during oestrus. Physical clearance is more efficient during oestrus (Evans, Hamer, Gason, & Irvine. 1987). Cellular responses are primarily phagocytosis. Efficient phagocytosis depends on a) mobilisation of an adequate number of neutrophils from the general circulation, with prompt migration of these cells through the endometrium and into the uterine lumen, b) adequate chemotaxis of neutrophils to contaminating bacteria, c) adherence of bacteria to the cellular membrane of the phagocyte (opsonisation) and d) ingestion and successful intracellular killing of bacteria by neutrophils. While all causes for failure of uterine defence mechanisms have not been identified, some significant influences are well known. Repeated and overwhelming infections or contaminations, with predisposing factors such as perineal abnormalities and pneumovagina, are undoubtedly related to reduction in efficiency of the mare's defence mechanisms. Reduced efficiency of neutrophils to phagocytise bacteria and a defect in opsonisation by complement and antibody have been identified in susceptible mares, (Liu, Cheung, Walsh, Miller, & Lindenberg. 1985; Asbury. 1984; Asbury, Halliwell, Foster, & Longino. 1980; Watson, Stokes, David, & Bourne. 1987; Watson, Stokes, & Bourne. 1987; Widders, Stokes, David, & Bourne. 1984) and this is not related to reduced or inadequate numbers of white blood cells (Liu, Cheung, Walsh, Miller, & Lindenberg. 1985; Liu, Cheung, Walsh, & Ayin. 1986). Failure of the uterus to mechanically evacuate contaminants and inflammatory products also appears to be very important. Myometrial activity may be reduced in older multiparous mares (Evans, Hamer, Gason, & Irvine. 1987) and physical clearance of non-antigenic markers from the uterus of susceptible mares was delayed when compared to resistant mares (Troedsson & Liu. 1991). If the cervix does not relax adequately or is compromised with adhesions or anatomical functional defects, evacuation may be impaired.
Clients are encouraged to have a full reproductive examination performed on their mares to help decide on the probability of successful resolution of the problem. Clients are sent a copy of the evaluation as soon as possible and are then encouraged to discuss the prognosis. It is preferable to perform the exam while the mare is in oestrus and definitely before the mare may enter anoestrus. The order of examination is always the same, and notes are written by an assistant at the time of examination to prevent accidental omission of data. Each year one or two mares presented will be pregnant, so invasive procedures such as culture and biopsy always follow rectal palpation and ultrasonographic examination. For daily exams, mares are herded into a long chute system and are teased or can wait for admission into the palpation shed. The system is designed to eliminate the need to catch any horse and is expedient and requires few personnel.
The prognosis for mares barren 3 years or more is always guarded provided they come from well managed breeding farms. Regardless of the number of years barren, we have more success in mares less than 18 years old. Breed is important with relation to ability to use AI or embryo transfer. Access to old breeding records is desirable and may suggest where and why the problems began and why previous treatments were not successful.
Body condition and estimated weight are recorded. Use of a scale or tape is preferable. Body condition, disease and chronic pain are related to timing of recrudescence to cyclicity. Hirsutism may be related to seasonal changes or a pituitary adenoma. Mares kept for an extended stay must always be sent home in better condition than when they arrived, although not overfat.
Height of pelvis relative to anal slope of vaginal lips, abnormalities of perineal body, clitoris or labia, presence or need to modify a Caslick, and propensity to aspirate air or pool urine are all noted with suggestions for methods of improvement and necessity.
Uterine tone and size and ovarian activity are assessed. Also abnormalities such as sacculation and pyometra are recorded. The main purpose of rectal examination is assessment of structure, tone and form. It is not possible to routinely determine the presence of uterine fluid with rectal palpation. Ovarian examination gives important information on cyclicity and ovarian abnormalities.
Quantity and quality of uterine fluid are related to degree of inflammation (McKinnon, Squires, Carnevale, et al. 1987). Uterine cysts are related to age and chronic endometritis (McKinnon, Squires, Carnevale, et al. 1987; Adams, Kastelic, Bergfelt, & Ginther. 1987). Abnormalities detected and undetected by rectal examination are visualised.
Manual examination is recommended. The vestibular sphincter and cervix are assessed for efficacy as barriers to infection. Cervical defects and adhesions are not uncommon. Many defects can be repaired surgically and some mares still can become pregnant despite absence of much of the cervix. If the cervix is relaxed, introduction of the whole hand into the uterus for manipulation is indicated. Small granulomatous lumps (2 to 4 mm) are occasionally identified and cysts and luminal adhesions can be removed or destroyed. Luminal adhesions imply a poor prognosis. Recognition of problems such as vesicovaginal reflux may on occasion only be made during oestrus when the reproductive tract is relaxed (McKinnon & Belden. 1988). However, some cervical defects are best recognised when the tone is increased as during dioestrus (McKinnon, Arnold, & Vasey. 1991).
Guarded culture techniques (Accu-Culshure-TM specimen collection and transport system, Pleasantville, New York 10570-292) from mares in oestrus are recommended (Waelchli, Kanzig, Gygax, Corboz, & Ruesch. 1993). Samples are cultured aerobically for 48 hours. Microaerophilic, anaerobic and fungal culture techniques are used when indicated (Waelchli, Kanzig, Gygax, Corboz, & Ruesch. 1993; Hinrichs, Cummings, Sertich, & Kenney. 1988; Wingfield Digby & Ricketts. 1982). Similarly, culture of uterine biopsy specimens may be necessary on occasion to detect causative agents of deep, chronic endometritis.
Results of uterine culture are best interpreted in relation to numbers of neutrophils detected from the endometrium by exfoliative cytology. However, mares with chronic endometritis may have little surface irritation (Roszel & Freeman. 1988).
A single biopsy may not always be totally representative of the uterus (Blanchard, Garcia, Kintner, & Kenney. 1987; Dybdal, Daels, Couto, Hughes, & Kennedy. 1991) but a sample from the corpus cornual junction (site of embryonic fixation) is one of the most accurate determinants of inflammatory conditions and cellular infiltrates. At the GVEH, cell types and changes are classified to give an individual grading for acute, chronic and fibrotic changes (0 to 3 for each). The biopsy interpretation is one of the most important determinants of subsequent foaling rates (Kenney. 1978) and, in addition, the degree of chronic versus acute inflammation helps decide therapy. We find that the grading system currently in use does not inform the veterinarian that referred the sample how to treat the mare. Others have suggested that the correlation to susceptibility to endometritis in the intermediate grading groups is not good (Troedsson, Demoraes, & Liu. 1993). The relative severity of acute, chronic and fibrotic changes with our system, enables us to explain to both the client and referring veterinarian which, if any treatments may be of benefit and what the expected results would be.
Examination for cytogenetic abnormalities and endoscopy for visualisation of pathological changes are performed as indicated.
MANAGEMENT PREBREEDING SEASON
Best results are achieved when we have the opportunity to assess and correct identified problems before the breeding season begins. Failure to treat an active chronic uterine infection during the non-breeding season is one more insult to an already compromised reproductive tract. The aim of prebreeding season management is to 1) eliminate uterine infection and inflammation, and 2) prevent further contamination prior to and during the next breeding season. Economic considerations are highlighted during this time when mares are repeatedly treated, clients receive accounts, and yet there are no signs of success because it is too early to begin breeding.
Correction of Physical Abnormalities
Many of the treatments during the non-breeding season require multiple vaginal invasions to penetrate the cervix, thus a Caslick operation or a breeding stitch may be damaged. Decisions on when to perform reconstructive surgery relative to treatment of infection or inflammation are generally obvious. Episioplasty and perineal body transection (Trotter & McKinnon. 1988; McKinnon, Arnold, & Vasey. 1991) are generally performed immediately unless active inflammation is present, because once healed, they provide little impedance to invasive treatment or examination procedures. Similarly it is pointless to begin therapy for mares whose major inflammatory problem is related to vesicovaginal reflux until the primary problem has been treated. Surgery of cervical problems has often not been attempted due to difficulties of exposure, failure to identify the problem and occasionally mares successfully carrying pregnancy despite cervical defects. We attempt repair of major defects if possible and find that strong caudal traction of the cervix with two pairs of cervical forceps, (Knowles forceps, Sontec Instruments, 6341 South Troy Circle, Englewood, CO 80111) in mares heavily sedated and given epidural or local anaesthesia, often enables good visualisation of and surgical access to the cervix in the caudal vagina or vestibule. Many mares will conceive and carry a foal to term despite having major cervical damage, so care should be used when advising on the breeding future of a mare with this problem. These mares frequently have difficulty establishing a pregnancy while they have a foal on them, but become better candidates for breeding the next season. Consequently many have a foal every other or every three years.
Treatment of Infection and Inflammation
Frequently we have no choice as to what stage of the cycle mares are treated. Many mares (~ 70%) have entered anoestrus during winter by the time they arrive for treatment. However, if possible it is best to restrict treatments during oestrus to use the effects of increased drainage, contraction (Evans, Hamer, Gason, & Irvine. 1987; Evans, Hamer, Gason, Graham, Asbury, & Irvine. 1986) and increased efficiency of phagocytosis (Washburn, Klesius, Ganjam, & Brown. 1982; Ganjam, McLeod, Klesius, et al. 1982; Asbury, Schultz, Klesius, Foster, & Washburn. 1982). Cervix adhesions are treated by local application of an oil based corticosteroid and antibiotic cream (R.R. Pascoe pers comm), and concurrent intrauterine elimination of infection. Treatment may involve daily topical application, disruption of adhesions and massage, for as long as 2 weeks. Although some mares will become pregnant by AI with adhesions of the cervix, these more frequently abort and the adhesions remain as a nidus of infection and thus should be eliminated. Uterine infection with concomitant large amounts of debris and fluid are treated by voluminous saline lavage, 1 litre at a time, until the effluent is grossly clear. Intrauterine antibiotics as determined by culture and sensitivity testing are administered after each daily lavage. When culture results are negative and active chronic infiltrative endometritis is identified from biopsy, a different and perplexing problem exists. It may be inappropriate to treat these mares as they possibly have already eliminated the infection and remain as mares susceptible to endometritis. However, our approach has been to attempt to reduce the inflammatory response as many of these mares have small quantities of uterine fluid present. Therapy is prolonged (7 to 15 treatments) with administration of broad spectrum intrauterine antibiotics performed daily or every other day. Concurrently, systemic oestradiol esters are administered for 1 month. If uterine adhesions are present, then local corticosteroid therapy is instituted. There are currently no reports assessing the efficacy of these treatments and caution is advised. Immunosuppression from the corticosteroids, while reducing the lymphocytic infiltrate, may well result in establishment of an infection resistant to antibiotics. There are no oestrogen preparations approved for use in horses (Varner, Blanchard, & Brinsko. 1988). In addition, use of long term oestradiol therapy, although apparently beneficial on the local uterine environment, has in our experience and in the hands of others (Nishikawa. 1959), resulted in long term suppression of ovarian function. Some treated mares fail to cycle until well into the breeding season. Also, to date we know of no method to determine which mares will respond to oestradiol therapy with prolonged ovarian suppression. Regardless of the method of treatment, a reproductive evaluation is performed to assess efficacy at conclusion of therapy. At this time final reproductive surgeries (ie. Caslick) may be completed. Local treatments after a Caslick can be administered by passing an infusion pipette into the vestibule, then guiding it through the cervix rectally, similarly, but with more difficulty, to intrauterine treatment in the cow.
Unfortunately, not all mares arrive for assessment and treatment during the non-breeding season and of those that do, not all respond to therapy enough to enable breeding to begin without added treatments. The primary goal of the prebreeding assessment period is to have a uterine environment capable of supporting spermatozoa long enough so they can reach the oviduct in a condition capable of initiating fertilisation. It only takes a few hours for spermatozoa to pass through the uterotubule junction and remain relatively free from toxic products in the uterus. However, fluid inflammatory products, when mixed with spermatozoa, cause an immediate decline in spermatozoal motility that is proportional to the amount of inflammatory products (Squires, Barnes, Rowley, McKinnon, Pickett, & Shideler. 1989). Also, addition of uterine fluid of grades 1 to 2 (McKinnon, Squires, Carnevale, et al. 1987) to spermatozoa prior to breeding mares by AI resulted in a decreased embryo recovery (P < 0.05) (Squires, Barnes, Rowley, McKinnon, Pickett, & Shideler. 1989). Decline in spermatozoal motility "in vitro" can be arrested by addition of an extender to the uterine fluid prior to addition of spermatozoa (Squires, Barnes, Rowley, McKinnon, Pickett, & Shideler. 1989). However it appears that for best fertility, removal of inflammatory products from the uterus prior to introduction of spermatozoa is the most logical approach. To meet these objectives, mares are examined in early oestrus. Ultrasonographic identification of uterine fluid and culture and sensitivity results are used to determine whether uterine lavage, local antibiotics or a combination of both are necessary to obtain a uterus free from inflammatory products at the time of breeding. It is an inappropriate use of time and finances to breed mares destined to return to oestrus, thus mares with abnormal uterine fluid accumulations detected at the time of breeding are often recycled as soon as PGF2µ is capable of causing luteolysis.
If the uterine environment is prepared properly at the time of breeding, then our goal is to prevent contamination at, or immediately after breeding. Clearly, breeding contamination is more easily controlled with AI than with natural service. However, both techniques result in introduction of bacteria. Research in this area was originally reported by Kenney, Bergman, Cooper, & Morse. in 1975. They demonstrated that hygienically collected semen from "non-infected" stallions, contained numerous types of aerobic bacteria and fungi. The total number of aerobic microorganisms in each of eight ejaculations from five stallions collected, ranged from 0.09 to 36 million. However, addition of raw semen to nonfat dry milk seminal extenders containing either penicillin-streptomycin (1,500 IU/ml and 1500 m g/ml respectively) or gentamicin sulfate (1 mg/ml) resulted in no growth on any of the subcultures from treated samples, including zero time which was after about 5 minutes of exposure. There was heavy growth in all subcultures from raw semen (Kenney, Bergman, Cooper, & Morse. 1975). Aerobic bacteria commonly isolated from the urethra, semen and prepuce of stallions are E. coli and other coliforms, Pseudomonas aeruginosa, beta haemolytic and nonhaemolytic streptococci (S. zooepidemicus), Klebsiella sp, haemolytic and non-haemolytic staphylococci, Proteus sp, and Corynebacterium. Further experimentation has demonstrated effective elimination of bacteria without affecting motility using seminal extenders containing either penicillin-gentamicin or polymyxin B sulfate (1,000 IU/ml). Thus, it appears that addition of raw semen to appropriate antibiotic-containing seminal extenders is one method of ensuring minimal contamination at the time of breeding (Danek, Wisniewski, Krumrych, & Dabrowska. 1994; Vaillancourt, Guay, & Higgins. 1993; Padilla & Foote. 1991; Blanchard, Varner, Love, Hurtgen, Cummings, & Kenney. 1987; Arriola & Foote. 1982; Timoney, O'Reilly, Harrington, McCormack, & McArdle. 1979). It is important to remember that some antibiotics affect spermatozoal motility at high concentrations and may adversely affect fertility (Pickett, Squires, & McKinnon. 1987; Jasko, Bedford, Cook, Mumford, Squires, & Pickett. 1993).
If possible, mares are inseminated without disturbing reproductive surgeries such as the Caslick. The perineum is diligently cleaned and dried as previously described (Pickett, Squires, & McKinnon. 1987) and 500 x 106 progressively motile spermatozoa (PMS) mixed with appropriate antibiotic-containing extender are inseminated. Proper technique to ensure cleanliness of the stallion and collection equipment is important (Pickett, Squires, & McKinnon. 1987). To ensure the antibiotics have had adequate time to eliminate bacterial growth, it is best to allow at least 15 minutes at 370C prior to insemination. If a longer interval is required, extended semen may be cooled to 200C and stored for at least 12 hours (Francl, Amann, Squires, & Pickett. 1987) and often considerably longer at 40C (Pickett, Squires, & McKinnon. 1987; Douglas-Hamilton, Osol, Osol, Driscoll, & Noble. 1984). To reduce contaminating organisms to an absolute minimum, a method was devised (Kenney, Bergman, Cooper, & Morse. 1975) to "wash" spermatozoa by dilution with an antibiotic-containing extender, followed by centrifugation (300 G) to produce a "soft" pellet, decantation of the supernatant and resuspension of the resulting pellet in fresh, warm extender. This technique has the added advantage of removing much of the seminal plasma which reduces motility after prolonged incubation (Pickett, Squires, & McKinnon. 1987) with minimal damage to spermatozoa (Pickett, Sullivan, Byers, Pace, & Remmenga. 1975), however it is time-consuming and may not be necessary.
Regardless of whether AI or natural service is used to breed mares, if a Caslick has to be opened it should be immediately apposed after breeding. Temporary apposition may be achieved with Michelle clips, however mares often become irritated by their continual reinsertion. Another technique is placement of breeding stitches that allow the penis or forearm to penetrate the vagina without damaging the labial commissures. Unfortunately, in many instances they are not effective because the vaginal lips are often not joined far enough ventrally to provide a good barrier to pneumovagina. Effective management often involves immediate replacement of stitches into the vulvar lips post service. Mares and stallions to be bred by natural service should be well cleaned. It is wise to avoid strong disinfectants that may cause overgrowth of potentially pathogenic bacteria after prolonged use. A technique of minimising contamination by prebreeding infusion of 100 to 300 ml of antibiotic-enriched seminal extender has been described (Squires, Barnes, Rowley, McKinnon, Pickett, & Shideler. 1989). This technique has advantages, however caution should be advised. For maximum reproductive efficiency, 500 x 106 PMS should be deposited into the reproductive tract (Pickett, Squires, & McKinnon. 1987). Lower spermatozoal numbers are quite effective in highly fertile stallions (Pickett, Squires, & McKinnon. 1987). However this information was derived from AI with small volumes of semen or semen plus extender. Recent information has indicated that spermatozoal concentration or volume of inseminate may be important factors in fertility. When mares were bred with 250 x 106 PMS in 100 ml of extender, embryo recovery rate was significantly depressed (13.6%, P < 0.001) compared to mares bred with these same spermatozoal numbers from the same ejaculates in 10 ml of extender (70.6%) (Squires, Barnes, Rowley, McKinnon, Pickett, & Shideler. 1989). This finding becomes important when mares are bred to stallions naturally that, due to frequent breedings, may have low spermatozoal numbers in normal ejaculate volumes (30 to 150 ml). Our approach for mares bred by natural service is to use ultrasonographic detection of quality and quantity of fluid combined with culture and sensitivity results to determine optimum treatments. For instance, if mares have a large volume of fluid detected, then voluminous lavage of the uterus with a physiological solution such as Dulbecco's phosphate buffered saline that is not expected to be detrimental to spermatozoal survival is instituted immediately prior to breeding. Increased temperature (41-450C) of infused fluids seems to aid in evacuation of uterine contents by increasing uterine tone. This procedure is slightly irritating, however the aim is to clear the uterus of inflammatory products and enable spermatozoa to have a relatively safe passage into the oviduct before further inflammatory products are released. If small quantities of uterine fluid are detected, then intrauterine antibiotics are infused prebreeding (> 12 hours). If uterine fluid is detected at the time of scheduled breeding, depending on the type and volume of fluid, the mare is either recycled, a small volume of antibiotic containing extender (< 50 ml) is infused immediately prebreeding or oxytocin is administered intravenously and the mare re-examined in 1-2 hours (Neuwirth, LeBlanc, Mauragis, Klapstein, & Tran. 1995; LeBlanc, Neuwirth, Asbury, Tran, Mauragis, & Klapstein. 1994). When organising timing of breeding the problem mare, much effort is directed toward trying to breed only once, just prior to ovulation ( 30 and < 40 mm is detected and when endometrial folds are prominent (McKinnon, Squires, Carnevale, et al. 1987). Following this strict guideline for induction results in most mares ovulating between 36 to 48 hours after treatment (Voss. 1993). Recent developments have shown that other drugs such as GnRH analogues (McKinnon, Nobelius, Figueroa, Skidmore, Vasey, & Trigg. 1993; Harrison, Squires, & McKinnon. 1991; Meinert, Silva, Kroetz, et al. 1993) are capable of precise, timed ovulation and are now commercially available for routine induction of ovulation in cycling mares (McKinnon, Nobelius, Figueroa, Skidmore, Vasey, & Trigg. 1993; Jochle & Trigg. 1994). Use of GnRH is expected to be associated with less immunogenicity because of its smaller molecular weight, and could be used as a primary induction agent or between cycles when hCG was administered. Mares are bred regardless of the number of preovulatory follicles and multiple ovulations are actively encouraged. There are few effective, commercially available drugs to increase the number of ovulations per cycle. Follicle stimulating hormone, although effective, is very expensive for equine use and crude pituitary extracts are not commercially available. Recently, immunisation against recombinant bovine inhibin subunit has been demonstrated effective in increasing ovulation rates in mares (McKinnon, Brown, Pashen, Greenwood, & Vasey. 1992). Conception rates are proportional to ovulation rates (Squires, McKinnon, Carnevale, Morris, & Nett. 1987), thus treatment by immunisation against inhibin should improve pregnancy rates in normal and subfertile mares and in mares bred to subfertile stallions. With intensive reproductive management, multiple pregnancies when diagnosed early present little difficulty in reduction to a singleton (McKinnon, Voss, Squires, & Carnevale. 1993; Pascoe, Pascoe, Hughes, Stabenfeldt, & Kindahl. 1987).
The aim of therapies in the immediate period after breeding is to: a) reduce infection and inflammation to create a uterine environment capable of supporting pregnancy, and b) prevent further contamination. Spermatozoa are safely in the oviduct within 4 hours of breeding (Bader. 1982) and are protected from inflammatory products in the uterus and/or uterine treatments by the utero-tubule junction after this time (Brinsko, Varner, & Blanchard. 1991). The embryo will not be released into the uterus until around 6 days after ovulation, however because most intrauterine therapies have an attendant degree of inflammatory response, and to allow time for foreign material to be expelled or absorbed, no treatments are administered after day 3. The cervix begins to exhibit increasing tone and improves as a barrier to infection within 2 days of ovulation, although it remains more relaxed when inflammation of the reproductive tract is present. In addition, the corpus luteum remains resistant to PGF2µ released from local inflammatory responses until at least day 5 after ovulation. All mares are infused with 100 ml of plasma approximately 6 hours after breeding. Plasma contains complement and is thought to increase the efficiency of the mare's cellular uterine defence mechanisms (Asbury. 1984). Broad spectrum antibiotics may also be administered at this time. Approximately 12-24 hours after breeding, the uterine response to breeding and contamination is assessed by ultrasonography. If fluid is absent, then plasma and antibiotics (if indicated) are administered daily for 2 additional days. If small amounts of fluid are detected, the same treatment is applied after first lavaging the uterus with buffered saline (pH ~ 7.0). Saline is quite an irritant, especially with low pH (Pascoe, Stabenfeldt, Hughes, & Kindahl. 1989). Usually only 1 or 2 litres are necessary to remove inflammatory products. When large amounts of fluid are detected, the fluid is recultured and removed by voluminous lavage until returning fluid is free from debris. Oxytocin is added (20 to 40 IU/L) to flushing solutions and given intravenously. Oxytocin increases myometrial contractions in oestrogen dominated reproductive tracts (Jones, Fielden, & Carr. 1991; Liu, Troedsson, Williams, & Pascoe. 1991) and may aid in expulsion of material. Fluids warmed to above 400C are not used at this time because the oviduct with gametes or early embryo is in close apposition to the uterus. Increased body temperature has been demonstrated to increase embryonic mortality. Final reproductive surgeries are completed on day 4 after ovulation (ie. Caslick) and if further treatment is necessary, it is administered systemically. Some mares are maintained on systemic antibiotics and phenylbutazone until an early diagnosis of pregnancy is possible at day 11 to 13 after ovulation. Phenylbutazone (1 to 2 gm daily) has been shown to prevent irritant PGF2-alpha release in mares subjected to uterine biopsy in dioestrus (Ellsworth-Swihart, Archbald, Ingraham, & Godke. 1985). Its use is discontinued if inflammatory uterine fluid is detected after day 7 or when a negative pregnancy diagnosis is confirmed (day 14). Due to potential toxicity of chronic phenylbutazone administration (Snow, Douglas, Thompson, Parkins, & Holmes. 1981; Collins & Tyler. 1985), it is used sparingly and discontinued after a foetus is detected within a vesicle (day 20 to 22). Mares that continue to produce large amounts of uterine fluid after ovulation, for multiple cycles, despite intensive management, may be poor candidates for future reproductive performance. These mares often have mild acute and severe chronic endometritis with varying degrees of fibrosis. Although not commonly recommended, after discussion with the client, and with informed consent, irritant therapy may be used. Irritant therapy in the form of dilute disinfectants such as Lugol's iodine or chlorhexidine acetate have been advocated. Some clinicians use intrauterine infusion of kerosene as well. The response to irritant therapy is acute endometritis which initiates stimulation of mechanical cellular and humoral responses. However, extreme caution is advocated as individual sensitivity sometimes results in irreversible degenerative changes with scarring and uterine adhesions (Mather, Refsal, Gustafsson, Seguin, & Whitmore. 1979). Without having informed consent of the client, one could perhaps be susceptible to malpractice claims if severe reaction and sclerosis occur after use of irritant intrauterine therapy. A more logical approach is milder irritation with dilute povidone iodine, hypertonic saline or bacteria-free filtrates of streptococcal cultures (Couto & Hughes. 1985).
MANAGEMENT OF PREGNANCY AND PUERPERAL PERIOD
The incidence of early embryonic death and abortion is higher in mares that are difficult to get into foal. Pregnancy wastage is mediated by a) luteolysis from PGF2-alpha release from an inflamed endometrium, (Neely, Kindahl, Stabenfeldt, Edqvist, & Hughes. 1979) b) absence of a pregnancy-specific factor secreted by a viable conceptus that functions to maintain ovarian steroid production (McDowell, Sharp, Grubaugh, Thatcher, & Wilcox. 1988) and c) failure of a damaged or compromised uterus to adequately nourish or support foetal development (Kenney. 1978). Genetic abnormalities would not appear to be a major factor contributing to equine early embryonic death after day 7 (Romagnano, Richer, King, & Betteridge. 1987). Despite the absence of any experimental evidence that a primary deficiency of progesterone production is a significant cause of equine pregnancy loss, many mares throughout the world are treated with natural progesterone or synthetic progestogens. (Allen. 1984). There is much documented literature on doses needed to maintain pregnancy in mares ovariectomised between 20 and 80 days of pregnancy (Shideler, Squires, Voss, & Eikenberry. 1981; Hinrichs, Sertich, Cumings, & Kenney. 1985; McKinnon, Squires, Carnevale, & Hermenet. 1988). Progesterone, which is the only maternal hormone needed to maintain early pregnancy (Shideler, Squires, Voss, & Eikenberry. 1981; Holtan, Squires, Lapin, & Ginther. 1979) is primarily ovarian in origin until day 100-150. The placental unit begins to secrete progestins from around day 60, however they are not capable of maintaining pregnancy in ovariectomised mares until after day 80 (Holtan, Squires, Lapin, & Ginther. 1979). Currently it is difficult to decide when progestogen therapy is appropriate and then, after initiation, when and how to withdraw it. Exogenous progestogen therapy will maintain pregnancy when luteolysis is mediated by PGF2µ secretion from inadvertent iatrogenic administration (Kastelic, Adams, & Ginther. 1987), administration of powerful endotoxins (Daels, Starr, Kindahl, Fredriksson, Hughes, & Stabenfeldt. 1987) or failure of the conceptus to signal pregnancy associated with restriction of early conceptus mobility (McDowell, Sharp, Grubaugh, Thatcher, & Wilcox. 1988). In these cases, supplemental therapy can be withdrawn after secondary corpora lutea formation or after the placenta is the primary source of progestins, however therapy should begin prior to the expected decline in peripheral progesterone levels. When problem mares are diagnosed pregnant at the GVEH, further therapy is determined by the amount of inflammation and infection at the time of breeding, presence or absence of uterine fluid at the initial pregnancy diagnosis and previous history of early embryonic death or abortion. When uterine fluid is detected 1 or 2 days after ovulation, the incidence of early embryonic death is dramatically increased (McKinnon, Squires, Carnevale, et al. 1987). Similarly, luminal fluid with a positive early pregnancy diagnosis suggests a poor prognosis for embryonic survival (Adams, Kastelic, Bergfelt, & Ginther. 1987; McKinnon, Squires, Carnevale, et al. 1987). These mares are treated with appropriately selected antibiotics and phenylbutazone for 5 to 7 days, then re-evaluated. When pregnancy continues, antibiotics are administered for 5 consecutive days every month. As with many treatments of endometritis, there is no experimental data available to document the efficacy of such treatments. However, at the GVEH we have been using this approach for 8 years now and have little doubt that the number of mares maintaining pregnancy is much higher than those mares not treated. Please recognise that it may not be possible to actually provide valid significant scientific data from these studies. When a mare has a history of habitually aborting or experiencing EED, the decision of whether to supplement or not with either antibiotics or progestins or both, is largely at the discretion of the owners. Many owners, having invested so much time and money in getting the mare pregnant, will feel compelled to supplement regardless of lack of documented efficacy and our discussions. Failure of differnet types of progesterone therapy (many synthetic compunds) to maintain pregnancy in ovariectomised or prostaglandin treated mares has been an interesting observation (McKinnon, Figueroa, Nobelius, Hyland, & Vasey. 1993, McKinnon AO, Lescun TB, Walker JH, Vasey JR and Allen WR.: 2000) that has highlighted an even more important wastage in time, money and lack of specified effect.
Working with problem mares may be an enriching experience for the veterinarian. It certainly is an expensive experience for most clients. The lack of experimental evidence documenting efficacy of various intrauterine treatments is somewhat embarrassing and reflects the absence of a good experimental model for susceptibility to infection. However, the pathophysiology of various contributors to uterine defence in the mare appears to be gradually, albeit painfully, elucidated. The recent demonstration of enhanced uterine clearance associated with oxytocin administration (Leblanc, Neuwirth, Mauragis, Klapstein, & Tran. 1994; LeBlanc, Neuwirth, Asbury, Tran, Mauragis, & Klapstein. 1994) has been encouraging, especially to those veterinarians who have been advocating it for years. The next decade holds exciting promise.