Discussion

Introduction:
Recent surveys have indicated per cycle pregnancy rates exceeding
60% in well-managed, naturally-mated Thoroughbred mares (Allen et al. 2007). When using frozen-thawed semen in commercial sport-horse AI programmes, mean per cycle pregnancy rates also now regularly exceed 45% (Newcombe et al. 2011). In general, it can therefore be concluded that horses are fertile animals. Nevertheless, there are individual mares in which establishing pregnancy requires intensive veterinary management and/or is a considerable challenge, despite there being no obvious abnormalities that would be considered likely to impede fertility. Furthermore, even when pregnancy has been confirmed, there is no guarantee that it will be maintained to term. Indeed, approximately 15% of pregnancies detected at Day 12 - 15 subsequently fail; strikingly, more than half of these pregnancy losses occur in the period before establishment of the definitive (chorio-allantoic) placenta, i.e. before Day 42 of gestation. Frustratingly, in the majority of cases we are unable to explain why it was not possible to either establish or maintain pregnancy and we cannot therefore decide rationally what, if anything, can be done to improve the chances of success at subsequent attempts. This is largely because our understanding of how pregnancy is established and maintained during the pre- implantation period (first 42 days) is relatively superficial, but also because some potential underlying abnormalities are difficult to diagnose. On the other hand, recent clinical models of pregnancy loss and advances in genomic and proteomic technology have generated novel information about the complex developmental and maternal-fetal signalling events critical to the establishment of a normal pregnancy.

Failure to establish pregnancy:
While stallion subfertility, inadequate breeding management and persistent post breeding endometritis are undoubtedly major contributors to the failure to establish pregnancy, less easily diagnosed potential problems include oviductal occlusion, and chromosomal abnormalities either of one of the parent animals (Lear et al. 2008) or arising spontaneously in the zygote or early embryo (Rambags et al. 2005). There is also anecdotal evidence that some mare-stallion combinations are incompatible, although the reasons for this incompatibility are not known. Clinically, oviductal occlusion (either unilateral or bilateral) is probably more common than previously anticipated but is treatable (Allen et al. 2006), whereas karyotypic abnormalities of otherwise normal healthy horses, and true mare-stallion genetic incompatibility, are untreatable but likely to be rare. While techniques to diagnose chromosomal or specific genetic abnormalities in equine embryonic cells have been developed (Choi et al. 2010), they are in their infancy and require embryo biopsy, cryopreservation and transfer if they are to be applied in a clinical context. Nevertheless, these emerging technologies should provide valuable information about the types and incidences of embryonic abnormalities.

Failure to maintain pregnancy:
That embryonic loss is a common feature of the first 6 weeks of equine pregnancy is not surprising when one considers that this is a period during which pregnancy maintenance is dependent entirely on progesterone produced by the primary corpus luteum (CL), when nutrient and gaseous exchange takes place via primitive placental precursors with no stable attachment to the endometrium, and when a mind-bogglingly rapid series of complex developmental events must take place to transform an undifferentiated ball of cells (the morula) into a conceptus supporting a fetus with a functional circulation and precursors of all the major organs. Clearly, even apparently minor perturbations during this sensitive period could have far-reaching consequences, not only in terms of embryonic survival but also in terms of the health of resulting offspring. With respect to embryonic survival, recent studies have shown that asynchrony between the developing embryo and host endometrium (Wilsher et al. 2011) and a temporary fall in circulating progesterone concentrations after Day 18 of pregnancy (Betteridge et al. 2011) will trigger developmental failure. In addition, ongoing gene expression studies have begun to reveal the complexity of maternal-fetal cross-talk during early pregnancy, and it is hoped that they will yield valuable information about the biology of pregnancy maintenance and, in time, help indicate the deviations from the norm that are most likely to compromise the developing conceptus. This should not only improve our understanding of what could go wrong, but may also enable more precise selection of mares for breeding soundness and indicate whether, and which, therapeutic intervention(s) may be useful to prevent pregnancy failure.