Effect of Prostaglandin in Early Diestrus or Progesterone and Estradiol Administration on Equine FSH-Treated Donor Mare Embryo Recovery and Recipient Pregnancy Rate

Prostaglandin in early diestrus compared with progesterone and estradiol administration for 10 days followed by prostaglandin was not significantly different in terms of embryo recovery and recipient pregnancy rate in equine follicle-stimulating hormone (eFSH)-treated donor mares.

1. Introduction

Embryo transfer offers distinct advantages to mare owners who wish to maximize the number of foals from a particular mare and also may be useful for mares when carrying a foal to full term is neither desirable nor possible. Economically, there is a need to increase the success rate of embryo transfer. Successful embryo transfer relies on careful and accurate reproductive management of both the donor mares and the recipient mares.

Demand for a reliable means of predicting or regulating estrus and ovulation in the mare has increased with the use of assisted reproductive techniques, such as embryo transfer, artificial insemination, and other forms of appointment breeding [1]. Prostaglandins are the most widely used hormonal therapy in the horse industry. Prostaglandin F2α (PGF2α) and its analogs, when administered exogenously, are valuable tools for inducing regression of the corpus luteum (CL) and subsequent return to estrus [2]. Prostaglandin F2α (5 mg, SQ or IM) is effective only when a mature corpus luteum is present, and the mare will come into estrus 1 - 6 days later [3]. The time to estrus varies depending on the follicular dynamics of the ovary. This range of response to PGF2α administration is problematic when donor and recipient mare ovulation need to be closely aligned for embryo transfer. Therefore, it is common to have a ratio of two recipient mares to each donor mare using PGF2α for estrus induction. The use of daily injections with a combination of progesterone (150 mg, IM, q 24 h) and estradiol 17η (10 mg, IM, q 24 h) for 10 days followed by PGF2α injection (5 mg, SQ or IM) is another estrus synchronization method commonly used in equine embryo transfer programs. This hormonal treatment causes follicular suppression and regression, and a new follicular wave emerges when the treatment is stopped. The progesterone and estradiol treatment effectively aligns one and two follicular wave mares and allows the donor and recipient mare ovulation to be tightly synchronized. This method requires fewer transrectal examinations, and often only one recipient is treated per donor mare.

Currently single embryo recovery attempts are common in equine embryo transfer, with a 50% embryo recovery rate per ovulation quoted for commercial operations. This is because induction of multiple ovulations or superovulation has been elusive in the mare, and superovulatory treatments are less efficient in mares than in other domestic species [4-7]. Equine pituitary extract has been used to stimulate multiple ovulations in mares but was not commercially available. Recently, a purified pituitary extract product, equine follicle-stimulating hormone (eFSH) has been studied and was reported to result in successful superovulation of donor mares [8]. There is limited information on the success of different estrus induction protocols combined with the administration of eFSH and no information on embryo quality and subsequent post-transfer pregnancy rate.

The main objectives of this study were to investigate and compare ovulation rates, embryo recovery, embryo quality, and subsequent pregnancy rate using two estrus synchronization methods combined with eFSH treatment in an embryo transfer program.

2. Materials and Methods

There were 12 donor mares and 36 recipient mares. There were 12 mare cycles per group, using a randomized crossover design, and each donor mare was used in 2 consecutive estrus cycles. Donor mares were examined daily during estrus using transrectal palpation and ultrasonography with a 5-mHz rectal probe, to determine the day of ovulation, and for 2 days after ovulation. Donor mares with pre- or post-breeding uterine inflammation were treated daily using saline lavage for up to 3 days after ovulation. Group 1 was progesterone and estradiol 17β (PE; 150 mg and 10 mg respectively, IM, q 24 h) for 10 days followed by PGF2α injection (5 mg, SQ or IM [a]) on the last day of PE treatment. Group 2 was PGF2α on day 5 after ovulation. In both groups, when a follicle >20 mm in diameter was detected in a donor mare, eFSH [b] was administered (12.5 mg, IM, q 12 h) until one or more follicles >35 mm in diameter was present. Donor mares were given human chorionic gonadotropin (hCG) [c] (2000 IU, IM) when a follicle reached 35 mm, and they were bred at 48-h intervals using artificial insemination (AI) with fresh semen or natural cover by one of two stallions of proven fertility until ovulation (day 0 = day of first ovulation). A minimum insemination dose of 100 million normal and motile spermatozoa extended 1:1 with Kenney [d] extender was used. Seven or 8 days after ovulation, the donor mares were flushed for embryos. All mares were given PGF2α (5 mg, SQ) on the day of embryo collection. An average of three recipient mares was synchronized for each donor mare.

In group 1, donor and recipient mares were treated with PE. In group 2, donor and recipient mares were treated with prostaglandin to induce estrus. Recipient mares were also administered hCG (2000 IU, IM) as needed to induce ovulation near the time of the donor. Donor mares that were randomly assigned to be administered PE first were followed until the first ovulation after their embryo recovery flush, and then 5 days later were administered PGF2α for the group 2 protocol. Mares that were assigned to group 2 first were administered PE 1 or 2 days after their embryo transfer flush.

AB Technology Vigro Complete Equine Embryo Flush Media (6 l) [e] were used for embryo recovery. Mares were treated with 40 IU oxytocin [f] IM after 4 l of solution had been recovered. Embryos were identified using a stereomicroscope, enumerated, and scored for quality. Poor embryos were given a score of 1, fair a score of 2, good a score of 3, and excellent embryos received a score of 4, using the morphologic criteria described by McKinnon and Squires [9]; however, the scoring system used had embryo scores increasing as embryo quality improved, which is opposite to their report. Vigro Embryo Holding Media [g] were used to rinse and hold the embryos before transfer. Embryos were held at room temperature and transferred within 1 h of recovery. Embryos were aspirated into sterile 0.5-ml straws with 0.25 ml of holding media and loaded into a flexible Universal pipette [h]. Embryos were transferred using a non-surgical, transcervical approach into the uterine body of the recipients. Recipients were used that had an ovulation that occurred in the range of 2 days before or after the donor mare. Recipient mares were evaluated for pregnancy status on day 14 after ovulation using transrectal ultrasonography. Results were analyzed using χ [2] and ANOVA.

3. Results

Twelve group 1 mares completed the PE treatment protocol. One mare had a large 50-mm follicle at the end of the 10-day PE treatment period. She was not given eFSH before breeding, and therefore her data were excluded from the analysis. Data are summarized in Table 1. In group 1, the mean duration of eFSH treatment was 7.7 ± 4.9 days (range, 2 - 18 days). Mares averaged 1.9 ± 1.1 follicles >35 mm with hCG. There was only one mare that had additional ovulations >24 h after the first, and she ovulated on days 0, 3, and 5. Two mares did not respond to hCG in 72 h. One mare had not ovulated by 9 days after hCG and was given an additional PGF2α injection (5 mg, SQ or IM). She ovulated one follicle 24 h later. Ten of 11 mares (90. 9%) administered PE ovulated from 9 to 12 days after PGF2α, and one mare ovulated 17 days after PGF2α. There were 18 ovulations in group 1, the mean number of ovulations was 1.6 ± 0.8, and the range was 1 - 3 ovulations per mare. Eleven embryos were recovered in this group. Mean embryo score was 2.4/4. Five of the 11 transferred embryos resulted in pregnancy in the recipient mares; therefore, the pregnancy rate per transferred embryo was 45%. Pregnancies were established from two fair and three good-quality embryos. No pregnancy arose from poor-quality embryos. Despite the flushes being performed on day 7 or 8, we observed two small for date embryos (day 6 poor embryos). One of these had a rent in the embryonic capsule; the other had the embryonic mass shrink away from the embryonic capsule.

Twelve mares in group 2 completed the prostaglandin treatment protocol. Data are summarized in Table 1. The mean duration of eFSH treatment was 8.4 ± 3.8 days (range, 4 - 16 days). Mares averaged 2.6 ± 1.6 follicles >35 mm with hCG. Two mares ovulated before hCG was given. One mare grew a large group of follicles and failed to ovulate. Three mares did not ovulate within 72 h of hCG. One mare had not ovulated by 4 days after hCG, so PGF2α was administered, and she ovulated 7 days after hCG. Therefore, for this group, the PGF2α to ovulation intervals were one mare on days 4, 6, 9, 11, and 12 and three mares each on days 8 and 10. Most of the mares with multiple ovulations had additional ovulations occur within 24 h. However, one mare had ovulations on days 0, 1, and 3. There were 30 ovulations in group 2, the mean number of ovulations was 2.5 ± 1.5, and the range was 0 - 5 ovulations. Fifteen embryos were recovered in this group. Mean embryo score was 2.2/4. Four of those 15 embryos resulted in pregnancy in the recipient mares; therefore, the pregnancy rate per transferred embryo was 26.7%. Pregnancies were established from two fair, one good, and one excellent-quality embryos. No pregnancy arose from poor-quality embryos. Flushes were performed on day 7 or 8, but we observed some small for date embryos (day 6 size embryos on day 7 or 8 flushes). A few of these embryos had a malformed, wrinkled embryonic capsule.

The combined PE and PG results were 48 ovulations, 26 embryos, and 9 pregnancies. There were 26/48 embryos per ovulation (54.2%), 9/26 recipient pregnancies per transferred embryos (34.6%), and 9/23 pregnancies per cycle (39.1%). There were 17/23 (73.9%) flushes where embryos were recovered, and 6 (26%) flushes where an embryo was not recovered, including the one cycle without ovulation; the overall percent of failed embryo recovery was 7/24 (29.2%) cycles.

4. Discussion

To our knowledge there have been no previous studies comparing embryo recovery or pregnancy rate using these estrus synchronization methods during eFSH treatment of donor mares in an embryo transfer program. We found few statistical differences between treatments. The PE-treated mares had a longer interval from PG to ovulation compared with the PG-treated mares (10.8 ± 2.3 days compared with 8.7 ± 2.3 days; p = 0.0456). The interval from PGF2α to ovulation for the main population of mares was 2.1 days longer; it was much more synchronous in the PE-treated mares than the PGF2α-treated mares, in agreement with previous reports [10].

The mean number of ovulations tended to be higher in the PG-treated group (p = 0.1062). In terms of total numbers, there were 30 PG ovulations and 18 PE ovulations. There were 15 embryos for 30 ovulations in the PG group, and 4/15 (26.7%) resulted in pregnancies. There were 11 embryos for 18 ovulations and 5/11 pregnancies (45.5%) in the PE group. These rates were not significantly different. The higher number of ovulations in the early diestrus PG treatment group might be a result of the recruitment or rescue of medium-sized follicles as well as small follicles and result in more overall follicular development. We had a number of mares in the PG group that had medium-sized follicles on day 5 after ovulation when PG was given. The trend (p = 0.1062) was for more ovulations to occur in PG- versus PE-treated mares (2.5 versus 1.6 ovulations, respectively). In PG-treated mares, there are more follicles available for stimulation. Two mares in this group ovulated five follicles, whereas three ovulations was the maximum number obtained in the PE-treated group. The PE treatment in most mares resulted in an ovary devoid of medium- to large-sized follicles. There were in essence fewer follicles present available for eFSH to stimulate. The follicles that emerged after PE was stopped represented a pool of young follicles. These follicles are likely to contain fresh oocytes that have a good potential for fertilization. The PE treatment may provide an advantage over the PG-treated donor mares if eFSH-induced superstimulation results in a superstimulated hormonal environment, because fewer pre-ovulatory follicles might result in less hormone secretion than in PGF2α-treated mares. The hormonal environment induced by eFSH requires further study. While there were no significant differences in embryo quality, median embryo quality score was higher in the PE-treated mares. There were five good to excellent, seven fair, and three poor embryos from the PG-treated mares and six good to excellent, two fair, and three poor from the PE-treated mares. The small for date embryos in both groups had morphological disturbances and therefore were graded as poor-quality embryos. We believe that embryonic development was arrested, and the embryos were devitalized. The misshapen, wrinkled embryonic capsule may indicate a loss of embryonic vitality. No pregnancies arose from poor-quality embryos in either group.

The embryo recovery rates per mare in our study (PE, 1.0; PG, 1.4) were lower than reported by Niswender et al. (1.9) [8] Squires et al. [11] evaluated a number of eFSH protocols and reported embryo recovery rates from 1.3 to 2.6 embryos per mare. Niswender et al. reported significantly higher pregnancy rates (1.8 pregnancies per mare) in five eFSH hCG-treated mares compared with 0.6 pregnancies per mare in 29 untreated mares, suggesting that embryo quality was not a significant factor in their eFSH hCG-treated mares. There was no untreated control group in our study; therefore, the underlying basis for the lower than expected pregnancy rate per embryo transferred cannot be determined. It is possible that the lower pregnancy rate is related to donor mare or recipient mare quality, technique, or effects of hormonal treatment. Treatment with FSH for ovarian superstimulation has been shown to impact embryo quality and quantity in other species [12].

We observed five mares that obtained more than three follicles >35 mm in diameter and did not ovulate in response to hCG within 72 h. One of the features of eFSH use that is reported to be problematic is the proportion of mares (10%) that excessively superstimulate and do not ovulate [i]. The reason for failed ovulation induction in these mares has not yet been determined. We observed follicles that continued to grow after the eFSH had stopped and remained static in two mares, one of which ovulated in response to a subsequent PGF2α injection. It may be useful to begin treatment with eFSH earlier (such as when follicles are 15 mm in diameter) to limit the number of eFSH injections, or similar to some cattle protocols, to stop eFSH treatment earlier (such as when follicles are 30-35 mm in diameter), and then induce ovulation 1.5 - 2 days later. It may also be possible to use lower doses of hormone. If a PG protocol is used, the treatment may need to begin closer to ovulation (day 3) or the injection of PG delayed to day 7. During the natural cycle, the primary follicular wave begins around day 12 in a progesterone-dominated environment, and luteolysis occurs 4 days later, near day 16. The eFSH could be given on days 3 - 7 with PG on day 5 or days 5 - 9 with PG on day 7. The number of eFSH treatments could also be modified based on follicular size. Administration of eFSH should be initiated before follicular dominance, which is achieved by <25 mm. One source of variability is the range in follicular size present in early diestrus. The currently recommended protocol from Bioniche Life Sciences includes discontinuing FSH treatment for 36 h before ovulation induction with hCG [j]. This approach has been used with success in the bovine embryo transfer and is referred to as "coasting".

The hormonal environment created by large groups of follicles from eFSH treatment requires further study. It is possible that high levels of estrogen or progesterone from large numbers of follicles may not be optimal for oocyte maturation, embryonic transport, or development. Other studies have shown that cycling mares treated with PG in early diestrus have lower levels of estradiol in the induced heat. hCG decreases the estrus period from 6 to 2 days. The combined effects of PGF2α, eFSH, and hCG on estradiol levels and subsequent pregnancy rate are not known. Lower estradiol levels may be an advantage if there are more follicles secreting estrogen.

The PE and PG treatment protocols are commonly used for equine embryo transfer. The PG protocol has lower drug costs than the PE protocol. The PE is currently only available through compounding pharmacies. In comparison, the PE protocol requires fewer transrectal examinations because the synchronization results are more predictable and reliable than the PG protocol, but daily handling is required to administer the PE injections. In the PE group, only one mare did not respond to the treatment by follicular suppression and follicular wave emergence, and two mares did not respond to hCG within 72 h of administration. In the PG group, two mares ovulated before hCG was given, three mares did not ovulate within 72 h, and one mare failed to ovulate. Therefore, the PE protocol might provide closer alignment between donor and recipient mares and may allow the use of fewer recipient mares to be synchronized per donor and substantially reduce the total cost of the embryo transfer program.

Economically the equine industry needs pharmaceutical agents to help increase the success rates with embryo transfer. In general, the overall success rate in terms of pregnancies from 23 embryo transfer procedures with eFSH was slightly higher than what would have been expected if the hormone had not been used. Commercially, a 50% embryo recovery rate and 50% embryo survival rate is commonly reported for an overall success rate in terms of pregnancies per cycle of 25% [1,13-16]. In this study, we achieved a 34.6% pregnancy rate per transferred embryo, and 9/23 (39.1%) pregnancies per cycle. This lower pregnancy rate per transferred embryo is probably related to poorer than expected embryo quality and may be related directly or indirectly to some of the following factors:donor/recipient selection/management, excessive ovarian stimulation, delayed donor ovulations, or carryover effects from sequential eFSH administration over two cycles. In the development of eFSH protocols, the cost-benefit ratio should be evaluated to determine the effect on reproductive efficiency. In this study, the trend appeared to be that, while there were fewer ovulations, more embryos developed from the progesterone- and estradiol-treated mares (45.5%) compared with prostaglandin-treated mares (26.7%). Fine-tuning of the eFSH treatment protocol to identify and characterize the changes in the hormonal environment, oocyte quality, and responsiveness to hCG are all areas that require future study. This information will aid our understanding of the superovulatory effects of eFSH in mares and may be applied to develop more efficient superovulatory protocols.

We thank the Alberta Agriculture Research Institute and Calgary Stampede Corporation, Calgary, Alberta, for financial support and use of horses, Mike Thomson for help with the horses, and Bioniche Life Sciences and UpJohn Pharmacia for support.

Footnotes

1. Lutalyse; Pharmacia Animal Health, Orangeville, Ontario L9W 3T3, Canada.
2. eFSH; Bioniche Animal Health, Belleville, Ontario K8N 1E2, Canada.
3. Chorulon; Intervet Canada Ltd. , Whitby, Ontario L1N 9T5, Canada.
4. EZ-Mixin; Animal Reproduction Systems, Chino, CA 91710.
5. Vigro Equine Complete Embryo Flush Media; AB Technologies, Pullman, WA 99163.
6. Oxytocin; Austin Division of/de Vetoquinol NA, Inc., Lavaltrie, Quebec JOK 1H0, Canada.
7. Vigro Embryo Holding Media; AB Technologies, Pullman, WA 99163.
8. Universal Pipette; Minitube, Ingersoll, Ontario N5C 3K1, Canada.
9. Squires E. Personal communication. 2005.
10. Squires E. Personal communication. 2005.