Relationships Between Utero-Ovarian Parameters and the Ovulatory Response to Human Chorionic Gonadotrophin (hCG) in Mares

A decrease in endometrial edema from the administration of human chorionic gonadotrophin (hCG) to 24 h later was significantly associated with subsequent ovulation within the 24- to 48-h window after hCG administration.

1. Introduction

Reproductive management of broodmares is aimed at coordinating breeding with the peri-ovulatory period. A single well-timed breeding per estrus is desirable to maximize reproductive efficiency. Ovulation induction is used to shorten estrus to minimize the number of breedings and to align ovulation with the fertile period.

Hormones such as gonadotropin-releasing hormone (GnRH) analogs, equine luteinizing hormone, and human chorionic gonadotrophin (hCG) are used to induce ovulation in mares. hCG is the most economical product. Response rates, as defined by ovulation within 48 h of administration of hCG, are reported to be as high as 70 - 90% [1]. The mean response time to hCG is 36 h, but there is a bell curve distribution of ovulation times, such that some mares ovulate within 24 h of treatment and others not until 48 h [1]. There are a proportion of mares (15 - 30%), that do not respond in a timely fashion to the administration of hCG, and presumably, ovulate on their own [1].

Ovulatory agents are particularly useful when breeding with semen that is limited in availability or by dose, such as with cooled transported or frozen semen. The fertile window for cooled semen is ≈48 h, whereas the fertile window of frozen semen in mares is reported to be ≈12 h [2]. Mares bred with frozen semen are frequently inseminated 24 - 28 h after hCG, or alternatively, are examined repeatedly at 6- to 12-h intervals until ovulation is detected and then bred [3]. Breeding at the time of ovulation detection requires frequent rectal and ultrasound examinations and increases the cost of breeding with frozen semen [3]. Pre-ovulatory breeding using fixed time insemination with frozen semen is preferable to many practitioners because it is efficient in terms of time management; however, two inseminations are performed [2].

To date, the presence of endometrial edema at the time of hCG administration has been proposed as one indicator of follicular maturation and responsiveness to hCG; however, others could not repeat this result [4,5]. Therefore, practitioners use time of hCG administration when planning to breed rather than objective criteria such as changes in uterine or ovarian parameters that indicate that hCG is working. There is a need to identify factors that may be used in an evidence-based approach after hCG administration to determine whether there is a high likelihood of ovulation, or conversely, if there is a high likelihood the mare will not ovulate in 48 h. If this were the case, the practitioner would continue to check the mare and follow her to ovulation, when using frozen semen, or if using cooled semen, they would evaluate the need to reorder.

The reliability or effectiveness if hCG after multiple administrations has been questioned when multiple injections are used in a single breeding season or over consecutive years in the same mare. It has been shown that the administration of hCG results in antibody formation in the mare; however, antihCG antibodies have not been shown to interfere with the biologic activity of the hCG product [6].

The objective of this study was to determine the effect of hCG (time 0) administration on ultrasonographic and rectal palpation features of the cervix, uterus, and ovaries that were associated with ovulation within 48 h of administration. A second objective was to evaluate a subgroup of mares that had not ovulated by 24 h after hCG and determine what changes in cervical, uterine, and ovarian parameters from time 0 - 24 h was associated with ovulation in the 24- to 48-h window. A meta-analysis of follicular measurements and uterine tissue ultrasound echotextural characteristics in mares that were administered hCG during estrus was used. Information from this study has practical application for veterinarians using assisted reproductive techniques or involved in intensive breeding management.

2. Materials and Methods

The experimental protocol was approved by the University of Saskatchewan’s Institutional Animal Care and Use Committee. A retrospective review of reproductive records from 353 light horse and draft cross mares that received hCG [a] during the 2002 - 2005 breeding seasons was conducted. Seven hundred seventy estrus cycles were reviewed for the study. Mares were maintained in dry lot pens and individually identified by numbered freeze brands or by unique head and body markings. Transrectal palpation and ultrasound were used to assess the ovarian and uterine parameters. Data were included in the study if the mare was examined by transrectal palpation and ultrasound at least once on the day of hCG administration and at least once daily thereafter after administration until ovulation or for 96 h after treatment. Mares were administered 2000 IU of hCG IM or IV. Follicular size (diameter of the largest follicle in mm), endometrial edema score (0, no edema homogenous echotexture; 1, light edema, not a spoke wheel pattern; 2, moderate spoke wheel edema; 3, prominent spoke wheel edema; 4, extremely prominent edema characterized by large black spaces in the endometrium) and uterine and cervical tone (soft, moderately soft, moderately toned, toned) at hCG administration were recorded. Because there were only two observations where the cervix was scored as toned, these data were combined with the moderately toned category. Follicular size and endometrial edema score at 24 h after hCG administration was recorded. The response to hCG (from time of hCGadministration to 24 h later) was evaluated in relationship to the following utero-ovarian parameters: diameter of the largest follicle (mm), uterine and cervical tone, follicular response at 24 h after hCG (follicle size increase, decrease, remain unchanged, or ovulated), degree of follicular response at 24 h after hCG (categorized in 5-mm increments that increased or decreased or unchanged or 0-mm change caused by ovulation), and edema response from time 0 to 24 h after hCG (increased, decreased, or unchanged). Some mares had or were receiving other concurrent hormonal therapies during the estrous cycle, such as prostaglandin; however, these were not analyzed statistically because of the large number of treatments. Season was divided as early (March-April), mid (May-July), and late (August-October).

An ultrasound machine [b] and 5-MHz linear array probe was used for all ultrasound examinations using standardized power and gain settings. Follicular size, the presence or absence of luteal tissue, and uterine and cervical tone were noted. The endometrial edema scores (range 0, no edema to 4, maximal edema) were recorded. A significance level of p <0.05 was used for all statistical tests.

Data were analyzed using logistic regression (PROC GENMOD SAS version 8.1) [c] to account for mares with multiple samplings. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. The OR for the dichotomous variables is reported as the odds of ovulation in one category compared with another, whereas with continuous variables, the OR is reported as the odds of the outcome for every 1 unit of increase of the continuous variable.

Two different logistic regression analyses were performed with two different dependent variables. The first dependent variable was ovulation in 48 h. The second dependent variable was ovulation between 24 and 48 h. The Bonferroni method was used to correct the p values in multiple post hoc comparisons.

3. Results

Descriptive Data

There were 770 estrus cycles included in the data set where ovulation in 48 h was the dependent variable. There were 103 cycles where mares ovulated sometime from time 0 to 24 h; therefore, these cycles were excluded from the second data set so that 667 cycles were included where ovulation from 24 to 48 h was the dependent variable.

The distribution of detected ovulations after hCG was as follows: 0 - 24 h, 13%; 24 - 48 h, 60%; 48 - 72 h, 13%; 72 - 96 h, 6%; >96 h, 8%. The proportion of mare cycles (n = 770) that had a follicle size (diameter of the largest follicle in millimeters) at hCG administration of <34 mm was 3.1%; 35 - 39 mm was 57.5%; 40 - 44 mm was 27.5%; 45 - 49 mm was 8.4%; 50 - 54 mm was 2.4%; and >55 mm was 1.1%.

The proportion of mare cycles (n = 636) with an increase in endometrial edema score from time 0 to 24 h after hCG was 17.0%, a decrease in edema score was 44.7%, and no change in edema score was 38.3%. The proportion of mare cycles (n = 692 with a cervical tone classification at hCG administration of toned to moderately toned was 22.5%, moderately soft was 63.3%, and soft was 14.2%. The percentage of mares ovulating within 48 h after hCG administration for consecutive hCG treatment cycles are presented in Table 1.

The influence of month of the year at the time of hCG administration and the percentage of mare cycles where ovulation occurred within 48 h are presented in Table 2.

The seasonal distribution and the percentage of mare cycles where ovulation occurred within 48 h of hCG administration are presented in Table 3.

Results of the Regression Analysis: Ovulation by ≤48 h

There was a significant effect of follicle size at hCG administration on ovulation by 48 h, with an OR 1.59 times (95% CI, 1.34 - 1.84) greater likelihood of ovulation for every 1-mm increase in follicle size (p = 0.0002). Significant differences existed in cervical grade at hCG(p = 012) in terms of response by 48 h. Mares with a toned to moderately toned cervix at hCG were 2.37 times (95% CI, 1.75 - 2.98) less likely to ovulate than mares with a soft cervix at hCG (p = 0.006). Mares with a toned to moderately toned cervix were 1.64 times (95% CI, 1.25 - 2.03) less likely to ovulate in ≤48 h compared with mares with a moderately soft cervix at hCG (p = 0.01). There was an overall effect of month of treatment on ovulation within 48 h of hCG treatment (p = 0.003). In general, mares treated from April to July had greater odds of ovulating in 48 h than mares treated in August and September. Results are shown in Table 4.

The number of hCG treatments per year (p <0.0001) had a significant effect on ovulation in <48 h. Mares were (OR, 1.52; 95% CI, 1.35 - 1.68) less likely to ovulate in response to hCG in 48 h for each additional hCG treatment per year. The cumulative number of hCG administrations over all years of the study also had a significant (p <0.0001) effect. Mares were (OR, 1.49; 95% CI, 1.34 - 1.63) less likely to ovulate with each additional treatment. Uterine edema score (p = 0.70), uterine tone score (p = 0.48), mare age (p = 0.98), season (p <0.25), and the interaction (p = 0.21) had no significant effect on response to hCG in ≤48 h.

Results of the Analysis: Ovulation From 24 to 48 h

There was a significant effect (p <0.0001) of edema response to hCG. Mares where endometrial edema increased were 3.25 times (95% CI, 2.74 - 3.76) less likely to ovulate from 24 to 48 h than mares where edema decreased by 24 h after hCG. Mares where the edema increased were less likely (OR, 1.74; 95% CI, 1.21 - 2.27) to ovulate from 24 to 48 h than mares that had their endometrial edema stay the same.

Overall season was a significant factor (p <0.001) in response of mares from 24 to 48 h after hCG. Mares in spring were 20.0 times (95% CI, 18.90 -  21.10) more likely to ovulate from 24 to 48 h than mares in the fall (p <0.001). Mares in the summer were 100 times (95% CI, 99 - 101) more likely to ovulate in 24 - 48 h compared with mares in the fall (p <0.001). Mares were less likely (OR, 3.71; 95% CI, 3.06 - 4.36) to respond to hCG with ovulation from 24 to 48 h in spring compared with summer (p <0.001).

Month of treatment with hCG (p <0.0003) had a significant effect on ovulation in the 24- to 48-h window (Table 2). The number of hCGtreatments per year was significant (p <0.001). Mares were (OR, 1.59; 95% CI, 1.42 - 1.76) less likely to ovulate with each increase in numerical number of treatments. The overall cumulative number of hCG treatments over the duration of the study was significant (p <0.0001; OR, 1.58; 95% CI, 1.43 - 1.73), with mares receiving more treatments being less likely to ovulate with increasing treatment number. Mare age, follicle response, follicle size, and the interactions were not significant.

4. Discussion

hCG is a glycoprotein hormone routinely used to induce ovulation in estrus mares. The hormone is economical and convenient to administer. In this study, the overall percentage of mare induced to ovulate within the 48-h period after hCG administration was 73.8% (568/770). It has been proposed that the ovulation response after hCG administration can be improved if other uterine parameters such as the presence of endometrial edema and a relaxed cervical tone are present at the time of hCG administration [7]. We observed no significant difference between endometrial edema score at the time of hCG administration and the ovulation response. This is contrary to previously reported findings. Samper [4] reported an ovulation response of >95% after hCG administration when a combination of endometrial edema score and follicle size were used as treatment criteria. We observed a significant difference between the endometrial edema score response at 24 h after hCG and ovulation within 48 h of hCG administration. The endometrial edema in a normal reproductively sound mare generally follows a physiological pattern. In early estrus, endometrial edema is low (endometrial edema score 1 - 2) as the estrus progresses, and the influence of estrogen increases, endometrial edema score increases. The endometrial edema will peak at a score of 3 - 4, which corresponds to the prominent spoke wheel pattern on ultrasound. As ovulation approaches, endometrial edema dramatically declines and in most mares resolves [4] Bragg et al [5]. reported that changes in endometrial edema occurred after hCG administration, but the time frame was compressed to 48 h. In our study, a decrease in uterine edema from time 0 to 24 h was significantly associated with ovulation within the 24- to 48-h interval from hCG administration.

McCue et al. [8] and Sullivan et al. [9] reported that there was a reduction in the efficacy of hCG after the second or third administration of the breeding season. Wilson et al. [10] also found similar reductions in the ovulatory response in estrus mares administered hCG during five estrous cycles. These findings have led some investigators to recommend that hCG be administered no more than twice to a single mare in a giving breeding season [1,11]. In this study, we observed that there was a significant decrease in the percentage of mares ovulating within 48 h of hCG administration after successive administrations; this was particularly evident after the second treatment. However, because most mares only received a few hCG treatments per year, the decrease in ovulation rate is not likely to significantly alter hCG use. We observed a significant difference in the percentage of mares ovulating within 48 h of hCG administration after successive administration of hCG over multiple breeding seasons. There is overlap in these two parameters; hence, this was expected.

It is important to note that we also observed a significant difference in the number of mares ovulating within 48 h of hCG administration by month of the year. There was no significant interaction between the month of the year in which hCG was administered and the number of hCG treatments. A difference was noted between ovulatory response and the season in which hCG was administered. This information suggests that ovarian responsiveness to hCG varies by season and is less effective early and late in the breeding season. There is a need to further evaluate whether repeated administrations of hCG results in a reduced ovulatory response or whether it is a function of season or month in which hCG is administered, because of the relatively few mare cycles in the data set from mares that had multiple treatments of hCG early or late in the year.

Barbacini et al. [2] stated that the ovulation distribution in the first half of the breeding season is clearly different from that seen in the second half. They suggested that the endogenous luteinizing hormone peak occurred earlier in mares during the later part of the season, which resulted in more mares having ovulations in the interval from 0 to 24 h after hCG administration. The authors concluded that practitioners should manage mares inseminated with frozen semen differently according to the time of the year and that an earlier mean time to ovulation occurred late in the season. We are in agreement with their findings and report that mares in March were 25 times more likely to ovulate from 24 to 48 h than mares in September.

An effect of increasing mare age on the ovulatory response to hCG has been previously reported [2,8]. Barbacini et al. [2] found that mares >16 yr of age had a wider ovulation distribution than younger mares (3 - 16 yr) and that the percentage of mares failing to ovulate within 48 h of hCG administration significantly increases with age. McCue et al. [8] reported that the percentage of mares failing to ovulate after hCG administration was higher in mares >15 yr of age. We did not find a significant effect of age; however, we did not have a high percentage of mares 15 yr of age.

In summary, hCG is effective in inducing ovulation in ≈73% of mares administered this drug. Follicular size and cervical tone at time of treatment with hCG are associated with ovulation within the next 48 h. A decrease in endometrial edema score from the initial administration of hCG to 24 h was associated with ovulation in the 24- to 48-h window. As the breeding season progresses, there is a reduction in the ovulatory response rate to hCG administration related to increasing number of treatments and seasonal factors. In the fall, fewer mares ovulate in the 24- to 48-h window. Practitioners may consider altering monitoring schedules and breeding guidelines based on the season, number of hCG treatments, and changes in endometrial edema.

The authors thank the Alberta Agriculture Research Institute and the Equine Health Research Fund of the Western College of Veterinary Medicine for financial support.

Footnotes

[a] Choralon, Intervet Canada Ltd., Whitby, Ontario, L1N 9T5, Canada.
[b] GE Ausonics, Universal Ultrasound, Bedford Hills, NY 10507.
[c] SAS Institute, Inc., Cary, NC 27513-2414.