Dr. Dietrich Volkmann
It is now common knowledge amongst the equestrian fraternity that hundreds of mares in central Kentucky aborted their late gestation pregnancies during late April and early May of this year. At the same time, an enormous number of mares that had conceived during February and the first half of March also lost their pregnancies. The syndrome was named Mare Reproductive Loss Syndrome or MRLS. The economic loss to the Thoroughbred breeding industry will run into many, many millions of dollars.
Apart from the urgent need to diagnose the cause of the MRLS (see Appendix A for the latest press release by the Gluck Center, coordinators of the investigation into the cause of the MRLS), there are some related questions and problems that require intensive and urgent investigation. On behalf of Cornell's College of Veterinary Medicine, first Dr D. Schlafer and later the PI (May 24-26, 2001), visited Kentucky in order to establish ways in which our College could make a contribution to the diagnostic investigations and research prompted by the abortion storm. Obviously, we all hope that the MRLS will never return to our breeding herds, but if it did and we had not gathered what information we could during the most recent outbreak we will start at the same zero point next time and get no further than we were before this spring.
During my visit to Kentucky I was made aware of the lack of specific investigations aimed at characterizing the hormonal profiles of mares that had lost their early pregnancies. Reports from stud veterinarians indicated that mares whose pregnancies had surpassed the first stage of forming a placenta and who had started forming endometrial cups (Day 35 after ovulation) at the time of the abortifacient insult were at much greater risk of losing their pregnancies than mares that had not reached that stage of gestation. This implies that only mares in which placentation had been initiated were susceptible to the MRLS and that most mares that had lost their pregnancies were now retaining their endometrial cups. The latter are glands that develop form fetal cells that migrate into the mare's endometrium (glandular lining of the uterus). The cells proliferate and form glands that secrete a hormone, called equine chorionic gonadotropin (eCG, formerly PMSG) (see Roser, 1999, for review on eCG1). These endometrial cups fulfill two functions: they secrete eCG and they aid in the immunological tolerance of the fetus inside the uterus of its dam. The eCG stimulates the development of accessory corpera lutea (CLs} by stimulating ovarian follicles to luteinize and secrete progesterone. About 60% of these CLs, form without ovulation, i.e. no eggs are released when these follicles become CLs (as is the case when estrous follicles ovulate and form CLs).
Between 120 and 150 days after conception the development of new CLs ceases, presumably, because eCG concentrations fall in response to the demise of the endometrial cups. By Day 180 of gestation the numbers of endometrial cups and accessory CLs, and progesterone levels all fall to zero and the pregnancy is maintained by 5α-pregnanes, secreted by the placenta from the 50th day to the end of pregnancy. The placenta also produces estrogens, some of which are unique to equine pregnancy. Estrone sulphate has long been recognized as an indicator of fetal viability during the latter two thirds of pregnancy. A third major hormone of placental origin is relaxin. All these placental hormones are synthesized by the placenta, but their precursors originate from the fetus. The endocrinology of equine pregnancy is described in great detail by Ginther2.
If a fetus dies between 40 and 140 days of gestation the endometrial cups (fetal tissue that is independent of fetal survival) continue to secrete eCG that continues to stimulate the development of anovulatory CLs on the mare's ovaries, preventing her from returning to normal, ovulatory cycles and being rebred during the same summer2. While this effect is a major component of the MRLS this year, it is a regular occurrence each year when smaller numbers of mares will suffer fetal losses during the endometrial cup phase and cannot be rebred.
ECG exerts its effect on the ovary by binding to the receptors of luteinizing hormone (LH). The affinity of eCG for these binding sites is lower than that of LH, but the release of LH from the pituitary gland is blocked by high progesterone levels during pregnancy. When a mare has lost her pregnancy, but not the CLs on the ovaries, the latter can be destroyed by the injection of one or more doses of prostaglandin F2α (PgF2α), allowing LH to be released from the pituitary. LH, however, cannot exert its usual maturation and ovulation inducing effects on the developing follicles on the ovary, because eCG will drive the same follicles to luteinize without ovulating.
This proposal encompasses two aspects of the early pregnancy loss in mares:
1. Through hormonal analysis of between 120 and 180 systematically collected blood samples from selected aborting and non-aborting mares in Kentucky, we aim to characterize the hormonal patterns that are associated with the pregnancy losses. This will confirm or reject our assumptions regarding the behavior of endometrial cups after the fetal losses, but it may also provide crucial information regarding the mechanisms leading to the pregnancy losses. Does the insult affect luteal function, endometrial function, placental function or does it primarily affect the fetus? Clinicians have observed that the earliest sign that a pregnancy was affected by the MRLS was the appearance of ultrasonographically detectable specs (echogenicities) inside the fluids surrounding the fetus. This was sometimes followed by detachment of portions of the fetal membranes from the endometrium and the very last event that signaled the end of the viable pregnancy was the loss of a fetal heartbeat. It appears then that placental development was impaired in affected pregnancies. We also know that not all pregnancies with abnormal fetal fluid eventually died. Using hormone production as an indicator of feto-placental well-being or disease, we may be able to predict whether placental function in affected individuals returned to normal or remained affected for longer periods of time, suggesting that fetuses that suffered, but survived the MRLS are likely to be either normal or abnormal at birth.
2. Rebreeding mares with endometrial cups is difficult (see above), but some mares in Kentucky were rebred, did ovulate, and did conceive during the period when we assume them to have had endometrial cups. We shall compare hormonal patterns in successfully rebred and unsuccessfully rebred mares. In addition, we shall try to identify the exogenous hormone treatment(s) (if any) that most successfully returned the mare to breedable status. Several methods (including destruction with a biopsy instrument and excision by laser surgery) have been employed by different veterinarians in their efforts to destroy the source of eCG, the endometrial cups, so that mares could be rebred. No practical method has, however, been found. The third aspect of our proposed work on the rebreeding of mares will thus be to test a novel approach to the manipulation of the effect of eCG on the ovaries by injecting mares that have endometrial cups with a highly specific (monoclonal) antibody against eCG (available from Intervet as a drug for use in cattle3-5). This should obviate the necessity to destroy the endometrial cups, while the temporary neutralization of circulating eCG will provide sufficient time for the mare to ovulate under the influence of her own LH. Concerns with this approach are: a) What will the effect of binding between 1-2 million IU of eCG be on the well-being of the mare? b) If the mare does ovulate and conceive, will she be able to maintain normal fetal development while she still has active endometrial cups from her initial pregnancy?