Prostaglandins and their analogues are commonly used in equine reproduction. Since the first reports of prostaglandin-induced luteolysis in the mare (Douglas and Ginther, 1972), their use has expanded from manipulation of the oestrous cycle to treatment of other conditions in broodmares (Table 1). Lack of knowledge may limit efficient use of these drugs as well as causing safety issues. The first part of this article series focuses on physiology, safety and prostaglandin use pertaining to the ovaries and oestrous cycle (anovulatory follicles, persistent corpus luteum, luteolysis and oestrous manipulation), while the second article outlines other clinical applications related to reproduction in the broodmare.
| Type of prostaglandins used | Reported uses of prostaglandins in broodmares | Notes on use and efficacy | Licensed or unlicensed |
|---|---|---|---|
| PGF2α | Luteolysis and oestrous cycle manipulation | Very effective if corpus lutea are mature (over 5 days old) (Douglas and Ginther, 1972; Oxender et al, 1975) | Licensed |
| Oestrus synchronisation | Not effective as the sole method of oestrus synchronisation (Bradecamp, 2007). Efficacy is much improved by combined use with progestins or progestin/oestradiol combinations (Blanchard et al, 1998) | Licensed and unlicensed depending on protocol and product | |
| Treatment of persistent corpus lutea | Very effective (Irvine, 1993; Coffman et al, 2016) | Licensed | |
| Treatment of haemorrhagic anovulatory follicles/luteinised unruptured follicles | Several reports of high efficacy (Coffman et al, 2014; Davolli et al, 2018; Crabtree, 2020). However, some studies have implicated PGF2α as a cause of these structures (Ginther and Al-Mamun, 2009; Cuervo-Arango and Newcombe, 2010a) | Unlicensed | |
| Termination of pregnancy | Effective in early gestation. As pregnancy progresses, multiple and higher doses are required (Van Leeuwen et al, 1983; Rathwell et al, 1987; Daels et al, 1995). Intracervical route has been reported to be very effective at 65 days gestation (Cuervo-Arango et al, 2015) | Licensed and unlicensed depending on protocol and product | |
| Foal fostering/treatment of foal rejection | Reports of high efficacy for both fostering (Daels et al, 2002) and treatment of foal rejection (Barker et al, 2019) | Unlicensed | |
| Treatment of uterine Infection | Some uterine infections themselves lead to the release of endogenous PGF2α release and luteolysis, causing the mare to return to oestrous earlier than expected (Stabenfeldt et al, 1976). If uterine fluid or infection are detected in dioestrous, exogenous PGF2α should be administered to bring the mare back into oestrus to allow appropriate diagnostics and management |
Licensed and unlicensed depending on protocol | |
| Uterine ecbolic | Uterine fluid is not always associated with infection and can be used purely as ecbolic. Potentially decreases pregnancy rates if used in postovulatory period (Troedsson et al, 2001; Brendemuehl, 2002). Uterine fluid accumulation is commonly encountered post mating/insemination and is not necessarily because of infection or endometritis (Canisso et al, 2020) | Unlicensed | |
| PGE1 or PGE2 | Treatment of oviductal blockage | Reports of high efficacy with appropriate case selection (Allen et al, 2006; Alvarenga and Segabinazzi, 2018). Does not increase pregnancy rates in reproductively normal mares (Donatsch et al, 2022) | Unlicensed |
| Cervical relaxation | Conflicting reports on efficacy (Le Blanc, 2006; McNaughten et al, 2014) | Unlicensed | |
| PGE2/PGF2α combination | Prevention of haemorrhagic anovulatory follicles/luteinised unruptured follicles | Reported highly effective with Flunixin meglumine induced haemorrhagic anovulatory follicles and luteinised unruptured follicles (Martínez-Boví and Cuervo-Arango, 2016) | Unlicensed |
Physiology
Prostaglandins are locally active lipid compounds with multiple hormone-like effects. Their name derives from the prostate gland, which was initially assumed to the be the source of prostaglandins (Von Euler, 1935). The prostaglandins' chemical structure gives rise to their names. The letter relates to the structure of the 5 carbon ring (eg prostaglandin F1 – PGF1), while the number relates to the number of double bonds in their hydrocarbon tail (PGF1 and prostaglandin F2 – PGF2).
Prostaglandin production is dependent on the activity of prostaglandin synthases (colloquially known as cyclooxygenases), which exist in 2 different isoforms (Smith et al, 2000). Cyclooxygenase 1 is expressed by most cells and plays a role in homeostatic functions. Cyclo-oxygenase 2 is induced by inflammatory stimuli, hormones and growth factors and is mostly responsible for prostaglandin production in inflammatory and proliferative or neoplastic processes (Dubois et al, 1998). Prostaglandins have effects on smooth muscle contraction, vasodilation, blood clotting and inflammation (Jones, 1972; Ricciotti and FitzGerald, 2011). Their effects differ depending on the type of prostaglandin, the tissue and the prostaglandin receptors involved. It is these effects that are responsible for undesirable side effects (eg colic, bronchoconstriction, abortion).
Prostaglandins and their effects are usually localised to their site of origin, but when released systemically they are metabolised rapidly in the lungs (Piper et al, 1970) (Figure 1). This is the case for cyclic equine luteolysis, where endometrial prostaglandin F2 alpha (PGF2α) is released into circulation, first passing through the lungs before reaching the corpus luteum, as there is no direct vascular connection between uterus and ovary (Ginther et al, 1972). It is reported that the equine corpus luteum is subsequently more sensitive to prostaglandins than in ruminants in view of limited bioavailability after lung passage (Douglas and Ginther, 1975; Kimball and Wyngarden, 1977; Ginther, 1992).
Licensed use of prostaglandins
Licensed equine veterinary medicines containing PGF2α are available in the UK for use in broodmares (Figure 2). However, there are no licensed veterinary products containing PGE1/2, so human medicines containing PGE1/2 must be used (under the cascade) and written informed consent for these should be obtained. Licensed indications do vary between the commercially available products so clinicians should be aware of these (by checking the products summary of product characteristics) and obtain informed consent where necessary. Important differences between the PGF2α products commercially available include:
PGF2α has been shown to be effective for luteolysis when administered by various doses and multiple routes (intramuscular, subcutaneously, intrauterine and intraluteal) (Douglas and Ginther, 1975; Weber et al, 2001). However, only the intramuscular route is licensed for commercially available products. One study also reported on the effectiveness of administration of PGF2α at an acupuncture point (Bai Hui) (Janini et al, 2023). It was reported that a lower luteolytic dose could be administered at this site, possibly reducing side effects, and was a viable alternative route to intramuscular administration.
Human safety considerations and side effects
Preparations of prostaglandins or their analogues can be absorbed via the skin or mucous membranes and there is a potential risk of serious side effects, particularly in some vulnerable groups (Lust et al, 2011). They can cause bronchoconstriction, and people with respiratory conditions or asthma should avoid contact or wear gloves when handling these drugs. In pregnant people, there is a risk of uterine contraction and induction of labour or miscarriage because of the effects of PGF2α and PGE1 (misoprostol is commonly used as an abortion-inducing drug in humans). Similarly, dinoprostone (PGE2) should be avoided by pregnant people, people with asthma, epilepsy, glaucoma, hypertension or compromised cardiovascular, hepatic or renal function.
Despite the theoretical risk, a literature search regarding adverse reactions to veterinary products containing prostaglandins revealed only one published report (Wilkins and Bowman, 1997). In this instance, an accidental needle stick injury of dinoprost tromethamine resulted in the miscarriage of a 15-week pregnancy in a female veterinarian. A request to the Veterinary Medicines Directorate was made to ‘obtain information of the incidence of adverse events in humans with respect to prostaglandins in veterinary medicines – in particular prostaglandins used in equine fertility work’. The Veterinary Medicines Directorate communicated that ‘the Veterinary Medicines Directorate data-base has been searched and we have found a reaction incidence of 0.00000120% for human adverse events following exposure to a prostaglandin veterinary medicinal product, across all species, for prostaglandin veterinary medicinal products indicated for equine use the reaction incidence was 0.0000006325%. In other words, for every 10 000 doses sold of a product containing prostaglandin, we have received <1 report involving a human adverse event, the overall incidence is considered very rare’ (Veterinary Medicines Directorate personal communication, 2023).
Although the reported incidence of these adverse events appears very low, the potentially serious nature of these side effects should be understood by clinicians and all possible steps taken to minimise exposure to people in vulnerable groups.
Patient safety considerations
Commonly seen side effects of exogenously administered PGF2α in equine patients are colic, sweating, trembling, diarrhoea, hypothermia and tachycardia (Irvine, 1993). It is worth noting that these effects have not been reported as a result of endogenously released uterine PGF2α. Prostaglandins can be grouped into naturally occurring (such as dinoprost and dinoprostone) or synthetic analogues (such as misoprostol or cloprostenol). The synthetic analogues of PGF2α have a luteolytic effect at a much lower dose and have been reported to cause comparatively minimal side effects than naturally occurring PGF2α (Alcántara et al, 2005).
Undesirable effects are usually transient and dose-dependent, starting within 20 minutes of administration and resolving within an hour (Irvine, 1993). No reports of long-term effects or fatal consequences have been reported (Goyings et al, 1977), although warnings of possible anaphylactic reactions are commonly seen in product datasheets.
Strategies to avoid or reduce PGF2α side-effects include:
Clinical applications related to oestrus cycle manipulations and conditions of the ovaries
Luteolysis and oestrous cycle manipulation
Luteolysis results in oestrus associated with typical behavioural signs. Exogenous PGF2α administration induces luteolysis in mares, shortens the luteal phase of the oestrous cycle and hastens the onset of the subsequent oestrus and ovulation (Table 2). The subsequent onset of behavioural oestrus and ovulation is dependent on the size of the dominant (largest) follicle present at the time of luteolysis (Loy et al, 1979; Lofstedt, 1988). These are general considerations for PGF2α-induced luteolysis in equine patients:
| Indication | Type of prostaglandin | Specific prostaglandin | Dose regime | Comments |
|---|---|---|---|---|
| Luteolysis and oestrous cycle manipulation (Luteolysis of a mature Corpus luteum 5 days post ovulation – a high proportion of normal mares are not responsive to a single dose of PGF2α until 5 days after ovulation; Allen and Rowson, 1973) | PGF2α | Cloprostenol |
|
Side effects are common |
| Dinoprost tromethamine | 1 x 5 mg intramuscularly | |||
| d-cloprostenol | Product dependent differences in licensed dose: |
Pregnancy rates have been reported to be lower in mares with shorter PGF2α–ovulation interval (Cuervo-Arango et al, 2015), but there was no negative effect on pregnancy rates in mares with dominant follicles <30 mm at prostaglandin administration and a treatment–ovulation interval of 8 days or more (Cuervo-Arango and Newcombe, 2010b).
Oestrus and ovulation synchronisation
Synchronisation of ovulation between embryo donors and recipients is important to optimise pregnancy rates (Oguri and Tsutsumi, 1974; Wilsher et al, 2010). It may also be desirable when there is a limited availability of a stallion, eg because of competition schedule. A longer follicular phase (and behavioural oestrus) means that ovulation synchronisation is more difficult to achieve compared to animals with shorter oestrous periods, such as cows (Lofstedt, 1988; Bradecamp, 2011).
Prostaglandins have been used to synchronise oestrus in mares using a protocol of two injections 14–15 days apart (Table 3). Although oestrus can be reliably synchronised, with 90% exhibiting oestrous signs 6 days after the second injection (Hyland and Bristol, 1979; Squires et al, 1981; Coffman and Pinto, 2016), the time of ovulation can vary, making it challenging to synchronise recipients and donors (Bristol, 1981; Coffman and Pinto, 2016). It has been reported that mares ovulate 7–10 days after the second injection (Bradecamp, 2007), and that 10 recipients would be needed for each donor in order to have an 80% chance of one mare ovulating within 24 hours of donor (Irvine, 1981). Therefore, prostaglandins on their own are unreliable for oestrus/ovulation synchronisation. Their practical use usually involves shortening of the luteal phase to induce oestrus, with additional use of ovulation induction agents to synchronise ovulation more closely between mares.
| Indication | Type of prostaglandin +/- other hormones | Dose regime | Published studies | Comments |
|---|---|---|---|---|
| Oestrus synchronisation | PGF2α | Luteolytic dose of PGF2α (Table 2), administered 14–15 days apart | Hyland and Bristol, 1979; Squires et al, 1981 |
|
| Progestins + PGF2α | Oral altrenogest (0.044 mg/kg bodyweight) for 10–14 days, followed by 5–10 mg Dinoprost tromethamine intramuscularly on the last day | Lofstedt and Patel, 1989 | Ovulation from time of prostaglandin administration is very variable and can range from 3–11 days | |
| Progestins + Oestradiol + PGF2α | Progesterone (150 mg) in oil and 17 ß – oestradiol (10 mg) injections intramuscularly for 10 days, followed by 5–10 mg Dinoprost tromethamine intramuscularly on the last day | Loy, 1980; Blanchard et al, 1998 |
|
A combination of prostaglandins with other hormonal treatments such as progestins and oestradiol allows tighter synchronisation of oestrus and ovulation in the mare (Loy, 1980; Lofstedt and Patel, 1989). Progestins have an inhibitory effect on luteinising hormone release from the anterior pituitary and suppress oestrous behaviour; however, they do not fully suppress follicular development (Garcia et al, 1979; Lofstedt and Patel, 1989). Oestradiol causes greater negative feedback on folliclestimulating hormone secretion (Burns and Douglas, 1981; Loy et al, 1982).
Synchronisation using progestins combined with prostaglandin has been studied but ovulation from time of prostaglandin administration is still very variable and can range from 3–11 days (Lofstedt and Patel, 1989). When progestins are combined with oestradiol, it leads to better oestrus/ovulation synchronisation between mares. This combination results in greater suppression of follicular development and, on discontinuation, a shorter window for synchronised ovulation (70% of mares ovulating between 10–12 days), particularly in combination with prostaglandins and ovulation induction agents (Blanchard et al, 1998). Unfortunately, there are no commercially licensed or compounded products containing oestradiol is currently available in the UK.
Treatment of persistent corpora lutea
The corpus luteum normally lasts 14–15 days in the non-pregnant mare (Ginther and Pierson, 1989) until endometrial PGF2α release causes luteolysis (Figure 1). Corpora lutea that fail to regress are considered pathological (Stabenfeldt et al, 1974). Potential causes of a persistent luteal phase are (McCue, 1998) (Figure 4):
This persistent luteal phase can last for 2–3 months and regardless of the cause, a single standard luteolytic dose of PGF2α will cause luteal regression and subsequent return to oestrous in 80% of cases (Irvine, 1993; Coffman et al, 2016). This response will depend on the maturity of the corpus luteum or anovulatory follicle at the time of PGF2α administration.
Anovulatory follicles
Anovulatory follicles occur when a dominant follicle of preovulatory size fails to ovulate (Table 4). The process by which this occurs can be classed as physiological, where a subordinate follicle regresses and another ovulates, or pathological, where a dominant or codominant follicle fails to ovulate (Crabtree, 2020).
| Indication | Type of prostaglandin | Dose regime | Published studies | Comments |
|---|---|---|---|---|
| Management to reduce likelihood of occurring | PGF2α | Avoid using in mares prone to producing these structures. Use the lowest dose possible that will cause luteolysis | Cuervo-Arango and Newcombe, 2010a | |
| Treatment to cause luteolysis once matured | PGF2α | A single luteolytic dose at 5–7 days post formation (Table 2) | Crabtree, 2020 | If a normal ovulation has occurred in addition to the anovulatory follicle – pregnancy could still result and therefore, treatment with prostaglandins would not be necessary |
| Treatment to cause destruction of haemorrhagic anovulatory follicles and luteinised anovulatory follicles leading to aluteal cycle | PGF2α | Twice daily injections day 0–3 followed by single daily injections day 4–5. Every injection contains twice the normal luteolytic dose (eg 10 mg dinoprost tromethamine) | Davolli et al, 2018 | Prevents anovulatory follicle formation, suppresses progesterone levels. Rapid return to subsequent oestrous and prevention of excessive large haemorrhagic anovulatory follicles which take longer to resolve |
| Treatment of preovulatory follicle to prevent formation | PGE2/PGF2α combination (Dinoprostone injectable solution or dinoprost tromethamine) | 1500 iu of HCG administered. 500 µg PGE2 or 125 µg PGF2α combination administered intrafollicularly 32 hours later | Martínez-Boví and Cuervo-Arango, 2015 | Resulted in 100% of mares ovulating normally (n=4) and 100% becoming pregnant if inseminated (n=3). Small number of mares used (n=4) and prevented luteinised unruptured follicles in induced cycles only. Promising if shown to prevent naturally occurring anovulatory follicles |
Pathological anovulatory follicles are also referred to as haemorrhagic anovulatory follicles, luteinised unruptured follicles or persistent anovulatory follicles. Instead of collapsing and releasing the oocyte and follicular fluid, they persist and either haemorrhage or persist without haemorrhage. The majority of anovulatory follicles (85%) luteinise, producing progesterone (haemorrhagic anovulatory follicles or luteinised unruptured follicles). The remaining persistent anovulatory follicles (15%) are not hormonally active and have little effect on the mares' reproductive cycle, so bear no clinical significance (McCue and Squires, 2002).
Anovulatory follicles are reported to occur in up to 8.2% of oestrous cycles (McCue and Squires, 2002) and are a common condition that clinicians may encounter. They can be difficult to detect if examinations are not frequent enough and can be mistaken for a normal corpus luteum with a central lacuna (Cuervo-Arango and Newcombe, 2013) or even granulosa cell tumours (Crabtree, 2011). For unknown reasons, some individual mares are prone to anovulatory follicles (which therefore decreases their fertility) with a reported rate of anovulatory cycles as high as 25–50% (Ginther et al, 2007; Cuervo-Arango and Newcombe, 2009).
Treatment regimens include:
Prevention of anovulatory follicles using a combination of PGE2 and PGF2α: Follicular wall rupture and ovulation are caused by collagenase enzymes induced by intrafollicular PGE2 and PGF2α (Reich et al, 1991; Sirois and Doré, 1997; Robker et al, 2000). Cyclo-oxygenase inhibitors have been shown to reliably block ovulation in mares by preventing production of PGE2 and PGF2α (Watson and Sertich, 1991; Martínez-Boví and Cuervo-Arango, 2016). In an experimental model of flunixin meglumine-induced anovulatory follicles (Martínez-Boví and Cuervo-Arango, 2016), intrafollicular injection of PGE2 and PGF2α (500 μg dinoprostone and 125 μg dinoprost tromethamine) blocked the effect of non-steroidal anti-inflammatory drugs. Anovulatory follicle formation was prevented in 4 mares (100% of those induced to form luteinised anovulatory follicles at their control cycle). Only three of these mares were inseminated at this ovulation, but all 3 were confirmed pregnant 14 days later. This procedure (if shown to be effective on naturally occurring anovulatory follicles) would be another option available to clinicians for mares prone to repeated anovulatory follicle formation. Systemic administration of PGF2α has also been trialled but failed to prevent flunixininduced anovulatory follicle formation (Cuervo-Arango, 2012).
Conclusions
Prostaglandins (PGF2α in particular) are commonly used in equine reproduction and have varied uses. They should be used judiciously while also shielding vulnerable groups (eg pregnant people, people with asthma). Although PGF2α very effectively causes luteolysis and is used mainly for this reason, there are concerns over its effects on subsequent pregnancy rates and anovulatory follicle formation. PGF2α alone is not efficient for oestrous/ovulation synchronisation in mares. Although PGF2α could be a factor in anovulatory follicle formation, it may also be an effective method of treatment.