In preparation for placement, the glass balls were sterilized by autoclaving. Initial attempts at autoclaving resulted in several broken balls. However, use of a liquid cycle with a temperature of 250°F (121°C) and pressure of 16 psi, with no prevacuum or dry cycle and a slow cool-down phase, did not result in further breakage.Upon entering the study, mares were monitored daily via transrectal palpation and ultrasonography for their progression through the estrous cycle. A glass ball was placed in the body of the uterus at the first examination following ovulation. The perineum was cleaned with cotton, tap water, and antiseptic dish detergent. A sterilized sleeve was donned and a small amount of sterile lubricant was applied to the back of the hand. Grasping the glass ball, it was manually carried into the vagina. The ball was placed in the cervical lumen and moved forward with the index finger to the caudal uterine body. After removing the hand from the vagina, the ball was located on transrectal palpation and pushed forward to the horn-body junction if it had not already moved to that position. Once the ball was positioned in the uterus, the vulva was again cleaned as previously described. The uterus was infused with 1 g of ticarcillin disodium (Ticar, SmithKline Beecham Pharmaceuticals, Philadelphia, Pa) in a 35-mL volume and each mare received 250 μg cloprostenol (Estrumate, Bayer Corporation, Shawnee Mission, Kan) intramuscularly to prevent a persistent endometritis if contaminants were introduced with the glass ball.At the end of the glass ball phase of the experiment for a mare, the ball was removed from the uterus during the following estrus when the cervix was softest. Occasionally a mare would require sedation to allow better manipulation of the ball per rectum. Mares with pendulous horns presented the most difficulty. Removal was accomplished by manipulating the glass ball, per rectum, caudally toward the cervix, through the cervix, and then to the vulva for retrieval. If the cervix was not fully dilated, a gloved hand was taken per vagina to the caudal cervical os and the glass ball was retrieved from the lumen.

Experimental protocol

This study was conducted at Auburn University in southeastern Alabama between May and October 2000. All mares used in this study had ovulated at least once in the season before being assigned to a treatment group. Twelve mares were randomly assigned to each treatment group (25-mm and 35-mm glass balls), representing 24 of 38 study mares. Control data were collected from 32 of 38 study mares during cycles in which no glass ball was in the uterus. Eighteen of the 32 mares used to collect control data were also treated with a glass ball during the study, whereas 14 mares were not treated. Control data were not available from 6 of 24 mares treated with a glass ball for management reasons beyond our control.An endometrial biopsy was taken from each mare during estrus in the cycle before being assigned to a treatment group, for comparison with a sample following removal of the glass ball. The follow-up samples were taken immediately after removal of the glass ball to ensure detection of inflammation, if present. An attempt was made to take the preplacement and postremoval samples from the same area near the uterine horn-body junction. Endometrial biopsies were evaluated as previously described.⁵ The evaluator was blinded to the group assignment and preplacement results of each mare.Throughout the study, mares treated with a glass ball were individually teased with a breeding stallion at a rail. Behavior was scored by response of a mare to the stallion using a categorical scale (0 = rejection, 1 = indifference, 2 = receptive) adapted from behavioral signs previously described.⁶Beginning on the day of glass ball placement, mares were examined daily by ultrasonography and palpation per rectum to monitor changes in the reproductive tract and ball location. The ultrasonic appearance of a glass ball in the uterus is shown in Fig 2.

Mares were also teased to a stallion, and blood was taken for determination of progesterone concentrations. Daily evaluation was continued until 2 ovulations had been detected; after this, daily blood sampling and teasing was continued until 2 more ovulations were detected. However, if at any time mares were found to maintain luteal tissue (continued ultrasonic evidence of a CL, palpable tone in the uterus and cervix, absence of estrus signs on ultrasonography, and absence of estrus behavior) for 35 days, daily evaluation was discontinued and the mares were moved to pasture and blood samples were taken weekly to monitor progesterone concentrations.In mares that did not experience prolonged luteal function (>35 days), the glass ball was removed from the uterus after 4 ovulations had been detected. In mares that experienced prolonged luteal function, the glass ball was removed after progesterone concentrations had fallen to <1 ng/mL followed by a subsequent rise to >4 ng/mL, indicating a subsequent ovulation. Mares that spontaneously expelled the glass ball were removed from the study upon discovery without further sampling.Immediately following removal of a glass ball, an endometrial biopsy was taken for comparison with the preplacement sample. Following removal of the glass balls, mares were bred during the next season under the protocol of another study. Standard breeding management for artificial insemination and several stallions were used. The results reported are for the season and reflect pregnancy outcome at 15 days after ovulation.Estrous cycles (n = 50) were observed in 32 of 38 study mares to establish an interovulatory interval and incidence of spontaneous persistence of the CL as a control for the effect of the treatment protocol. Observations were made during separate control cycles when a glass ball was not in the uterus of any mares that were also used in a treatment group during the study. At least one cycle was evaluated for each of the 32 mares, with some contributing a second cycle. The reproductive tract and circulating progesterone concentrations were evaluated in the same fashion and on the same daily schedule as the treatment groups. During the control cycles, the mares were simultaneously being observed to establish estrous cycle control data for another study. Therefore, when spontaneous persistence of a CL occurred and a luteal phase lasted 30 days, the mare was given prostaglandin to lyse the CL.

Progesterone assay

Circulating progesterone concentrations were used to reflect luteal function. Concentrations higher than 1 ng/mL were considered indicative of functional luteal tissue. Plasma was harvested from blood collected from each mare. Plasma samples were frozen and held at −50°C until assayed in batches of approximately 200. Circulating concentrations of progesterone were quantified using a commercial radioimmunoassay kit (COAT-A-COUNT progesterone radioimmunoassay kit, Diagnostic Products Corporation, Los Angeles, Calif).

Statistical analysis

Two measures were derived from each interovulatory period: the interovulatory interval in days and the number of days during which progesterone was >1 ng/mL. The effects of glass ball size, monitoring method, mare and their interactions were tested using the GLM procedure of Statistical Analysis System (SAS Institute, Cary, NC). Ages of mare among groups were compared with use of an unpaired t test (GraphPad InStat version 3.00 for Windows 95, GraphPad Software, San Diego, Calif). The proportion of mares experiencing extended luteal function or spontaneous persistence of a CL during the treatment and control cycles was determined. In addition, the proportion of diestrus periods that resulted in extended luteal function or spontaneous persistence of a CL was determined. The proportion of mares and diestrus periods in which extended luteal function occurred during treatment and control cycles were compared using a Fisher Exact Test (GraphPad InStat version 3.00 for Windows 95, GraphPad Software, San Diego, Calif).

Results

A very small amount of uterine fluid (<1 cm depth) was observed via ultrasonography in 3 mares for 2 days following placement of the glass ball. By day 3, however, the fluid was no longer visible in any of the mares, one of which did go on to maintain luteal function for an extended period. None of the mares that developed uterine fluid experienced spontaneous loss of the glass ball.A total of 24 mares had a glass ball of either 25 mm (n = 12) or 35 mm (n = 12) diameter placed in the uterus. The 25-mm glass ball was spontaneously expelled in 6 of 12 mares (50%). Five were expelled within 24 hours of placement and a sixth during a subsequent estrus period, 11 days following placement. None of the 35-mm glass balls was spontaneously expelled.The glass ball was observed to randomly alternate between the left and right uterine horn-body junctions. Movement was observed in every mare except two. In those 2 mares, the 35-mm glass ball was consistently observed at the same site during each examination. One of the mares experienced extended luteal function and the other did not.Overall, 7 of 18 mares (39%) that maintained the glass ball experienced extended (>35 day) luteal function. Extended luteal function was detected during the first diestrus after ball placement in 4 mares, during the second diestrus in one mare, and during the third diestrus in 2 mares. Mean (±SEM) progesterone concentrations during the extended luteal periods are reported in Figs 3, 4, and 5.

There was no difference between the 25- and 35-mm balls in terms of proportion of mares having extended luteal function (2 of 6, 33%, and 5 of 12, 42%; P > .05). Extended luteal function occurred in 7 of 62 diestrus periods (11%) among mares following ball placement. Again, there was no difference between 25- and 35-mm balls in terms of proportion of diestrus periods resulting in extended luteal function (2 of 20, 10%, and 5 of 42, 12%, P > .05).An extended luteal period occurred in 4 of 32 mares (13%) observed for control data. Progesterone concentrations remained above 1 ng/mL for 30 days after ovulation in 4 of 50 control diestrus periods (8.0%) observed. Of the 4 mares that experienced extended luteal function during the control cycle, a glass ball was placed in the uterus of 3 of the mares during the treatment cycles. However, none of the 3 mares experienced extended luteal function while the glass ball was in the uterus. The proportion of mares that experienced extended luteal function (7 of 18, 39%) while a glass ball was in the uterus was greater than the proportion of mares that experienced an extended luteal period (4 of 32, 13%) during the control cycle (P = .04). The proportion of diestrus periods in which extended luteal function occurred was the same whether a glass ball was present in the uterus (7 of 62, 11%) or not (4 of 50, 8%; P = .75).The interovulatory interval was 23.0 (±0.43) days for the control cycles (n = 46) in which an extended luteal period did not occur. This was longer than the interovulatory interval (20.2 ± 0.41 days) for the cycles (n = 55) that occurred subsequent to glass ball placement without apparent extension of luteal function (P < .001). A functional CL was maintained (15.5 ± 0.35 days, range 11 to 23 days) longer in control cycles than in cycles with glass balls (13.2 ± 0.42 days, range 7 to 18 days) in which extended luteal function was not apparent (P < .001). Mean (± SEM) progesterone concentrations for the control cycles (n = 46) and treatment cycles in which an extended luteal period did not occur are presented in Fig 6.

The mean age of all mares in the study was 9.6 years (range, 3 to 20 years). Mares that experienced extended luteal function were younger, at 8.3 years (±0.87), than mares that did not experience extended luteal function, at 12.6 years (±1.05, P = .012). Mares that spontaneously expelled the glass ball were younger, at 6.3 years (±2.0), than mares that did not expel the glass ball, at 10.9 years (±0.87, P = .024).The interassay and intra-assay coefficient of variation for the progesterone assay was 7% and 3%, respectively. The sensitivity of the assay was 0.02 ng/mL. Estrous behavior observed during the study accurately reflected circulating progesterone concentrations. All mares with a glass ball invariably displayed behavioral estrus scores of 1 or 2 when progesterone concentrations were <1 ng/mL, while scores of 0 were observed when concentrations were >1 ng/mL. Those that experienced extended luteal function also displayed scores of 0 throughout the period while progesterone concentrations were >1 ng/mL and estrus behavior was monitored. A single CL was observed at the primary ovulation site throughout the observation period. No additional CLs were observed in any mare with a glass ball following the primary ovulation.No change in endometrial category was observed between the preplacement and postremoval endometrial samples in any of the mares in the 25-mm glass ball group. In the 35-mm group, the score declined by a category in one mare, improved by a category in 2 mares, and was unchanged in the other 9 mares. The difference in category assigned in those 3 mares was attributed to mild changes, up or down, in the amount of lymphocytic inflammation observed. The mare that declined by one category did not experience extended luteal function, whereas 1 of 2 mares with improvement in endometrial category experienced an extended luteal period.During the season following glass ball removal, 23 of 24 mares were bred, including all of the mares that had experienced extended luteal function. During the season, 17 of 23 (74%) of the mares bred subsequently conceived, including 5 of 7 (71%) of those that had experienced extended luteal function.

Discussion

Placement of the glass ball through the cervix was relatively easy in most of the mares. Occasionally the cervix of a mare would require some degree of manual dilation to push a 35-mm diameter ball through its lumen; however, the 25-mm balls generally passed with ease. During preliminary work, we had found that the glass ball was more likely to be expelled from the uterus if it was placed a day or two before ovulation. Based on the anecdotal information from the Netherlands (message to Equine Clinicians Network, Dr Randy J. T. de Greef, March 19, 2000) and our preliminary experience, we decided to place the glass ball in the uterus at the examination following ovulation. Our assumption when placing the glass ball following ovulation was that the cervix would have started to close under the influence of rising progesterone, which might help prevent it from being expelled. We found that younger mares were more likely to expel the glass ball, perhaps because younger mares have more effective uterine clearance (in estrus and the periovulatory period) and a less dependent uterine position than do older mares. The smaller diameter and lighter weight of the 25-mm glass ball also may have contributed to the spontaneous expulsion from 6 mares, although it is also possible the cloprostenol contributed to the loss in some mares. However, losses observed during preliminary work were not associated with cloprostenol administration, and neither was the loss in this study from the mare that expelled the ball during the subsequent estrus at 11 days following placement.Idiopathic persistence of the primary CL, also known as spontaneous persistence of the CL,⁷ cannot be completely ruled out as an explanation for the extended luteal function observed in this study. The incidence of idiopathic persistence is reported to vary widely.^{8 and 9} Ginther and Pierson⁸ did not observe idiopathic persistence in any of 69 interovulatory intervals, while Stabenfeldt and Hughes⁹ suggest it can occur in as many as 25% of estrous cycles. However, acceptance of inadequate evidence for the condition may have led to an overestimation of the incidence in some reports.⁷The proportions of diestrus periods that resulted in an extended luteal period were not different between the control (4 of 50) and glass ball (7 of 62) cycles. However, the proportion of cycles in mares with glass balls inserted is heavily biased by the 4 cycles from each mare that did not experience extended luteal function. A greater proportion of mares (7 of 18) experienced extended luteal function when a glass ball was in the uterus than without a ball (4 of 32) during the control cycles. Four of the mares that experienced extended luteal function with a glass ball in the uterus did so during the first diestrus following placement and did not experience subsequent ovulations. These mares are consequently underrepresented in the proportion of diestrus periods among the mares that had a glass ball inserted, especially if all or some of the mares had experienced additional periods of extended luteal function following subsequent ovulations. This would seem to make the proportion of mares a more valuable indicator of glass ball efficacy. However, in all fairness, we should point out that 2 of 7 mares that experienced extended luteal function did so after the third ovulation following glass ball placement (Fig 5). The control data were collected from 32 mares in 50 estrous cycles; only 18 of the mares were observed in more than one cycle. This may have biased our results for fewer occurrences of idiopathic persistence of the CL, although we believe it is unlikely.Five of the 7 mares in this study that experienced extended luteal function when a glass ball was in the uterus had never previously been observed to have prolonged interovulatory intervals over multiple seasons. Historic data were not available for the other 2 mares. The average length of the luteal period reported to be associated with idiopathic persistence of the CL is approximately 2 months.⁷ The average length of luteal function observed in this study was 3 months. These several points would seem to support the idea that the extended luteal function observed in this study was indeed affected by the glass ball protocol rather than idiopathic persistence. Therefore, although it is not possible to positively distinguish idiopathic persistence of the primary CL from extended luteal function influenced by the glass ball protocol, we believe that our observations in this study suggest a genuine effect. An explanation is not readily apparent for our observation that mares experiencing extended luteal function were younger than those that did not experience extended luteal function.Two possible explanations for an effect of the glass ball have been discussed (Equine Clinicians Network archives). First, the glass ball simulates a conceptus and through movement and physical contact prevents prostaglandin release from the endometrium, in turn maintaining the CL (an endogenous progesterone source) indefinitely. A second theory suggested that the glass ball would stimulate mild inflammation, in turn causing release of small amounts of prostaglandin that would be inadequate to achieve luteolysis. As long as the glass ball was present, the endometrium would remain in a prostaglandin-depleted state and the CL would be maintained.In the event that the first theory discussed was correct, two glass ball sizes were evaluated in this study to account for a range in vesicle diameter that would be expected to occur naturally during the early stages of pregnancy. However, this theory seems to assume that the physical presence of a spherical structure, in this case a glass ball, in the uterine lumen will prevent prostaglandin release. This is contrary to conventional logic that would assume, as has been demonstrated in other species, that a chemical messenger is produced by the equine conceptus to allow maternal recognition of pregnancy and avoid prostaglandin release.¹⁰The second theory is as equally confusing, considering that there are countless reasons why a mare may have low-grade endometrial inflammation, yet clinically we do not recognize scores of mares that maintain their luteal tissue indefinitely. Both theories suggest that the glass ball is mobile enough to contact the majority of the endometrium or cause low-grade endometrial inflammation throughout the uterus. Our findings did not support the idea that the glass ball was particularly mobile in the uterus. Although the ball did move between the uterine horn-body junctions in most mares, the distance moved was only a few centimeters. In 2 mares, the glass ball did not move at all. One of the mares experienced extended luteal function in spite of the lack of ball movement. It probably also would be more logical to assume that an irritant to the endometrium would cause low-grade inflammation, which in turn would likely trigger sufficient release of endogenous prostaglandin F_2α to cause luteolysis.^{11 and 12} Uterine biopsy results did not reflect an increase in endometrial inflammation. Further, the interovulatory interval and functional life of the CL was >2 days longer during control cycles than in cycles when a glass ball was in the uterus. This would suggest that the ball was more likely to cause early regression of the CL.Placing the glass ball following ovulation could predispose a mare to endometritis, considering the procedure involved passing a foreign object, although sterilized, through the cervical lumen after a mare has entered diestrus. A mare susceptible to endometritis may not have time to clear her uterus of contaminants before closing the cervical lumen completely. Based on this rationale and experience during our preliminary work, we decided to provide treatments simultaneous to glass ball placement that were intended to help prevent a persistent postplacement endometritis. Hence, each mare in this study was infused with ticarcillin disodium (Ticar) and treated with cloprostenol (Estrumate) following glass ball placement. Uterine fluid was only observed in a few mares in the first few days following glass ball placement, and it resolved quickly. Nevertheless, it is advisable to re-examine a mare with ultrasonography following placement of a glass ball to ensure a detectable endometritis has not developed.Discussions on the Equine Clinicians Network suggested that the glass ball had no long-term detrimental effects on the uterus. The relatively minor ultrasonic changes detected in the uterus following glass ball placement, the endometrial biopsy results, and a conception in 74% of the mares following glass ball removal would seem to support this claim.It is interesting that the mares experiencing extended luteal function were able to maintain progesterone concentrations above 1 ng/mL for an average of nearly 3 months. This is the period in which we would expect the fetoplacental unit to begin assuming maintenance of pregnancy through the production of pregnanes in an ever-increasing number of pregnant mares. The progesterone profiles in these mares were very similar to those reported for mares hysterectomized 3 days following ovulation.¹³ The primary CL was present 70 days following ovulation in hysterectomized mares, but disappeared by 140 days.¹³ We found this was also true of mares that experienced extended luteal function with a glass ball in the uterus.The efficacy of an intrauterine glass ball for maintaining luteal function and thus preventing cycling and behavioral estrus in mares appears to be moderate. Our results were not quite as good as those reported from the Netherlands (Dr Randy J. T. de Greef, message to Equine Clinicians Network, March 19, 2000). The Dutch veterinarian indicated that the technique works in at least 75% of cases; in contrast, we found the technique to be effective in only approximately 40% of mares. Perhaps a placebo effect for mare owners, as is suspected with the use of progestin implants, would explain the additional success reported from the Netherlands.The glass ball protocol takes advantage of endogenous progesterone production to suppress behavioral estrus. Some variation in estrus behavior will be observed in any group of mares teased to a stallion throughout the cycle. However, mares are typically expected to reject a stallion when a functional CL is present and to change from indifferent to receptive as estrogens rise in the absence of a functional CL. Teasing results in this study were consistent with the behavior expected for the concentration of circulating progesterone detected. We did not monitor the mares that experienced extended luteal function beyond their subsequent ovulation, although we speculate that if the glass ball had been left in the uterus, some of the mares may have experienced another extended luteal period. Another researcher related information to us about 2 mares in which he had placed a glass ball (personal communication, Dr Peter Daels, National Institute of Agricultural Research, Nousilly, France, April 2000). The mares experienced extended luteal function, then, following administration of prostaglandin, both mares returned to estrus, retained the glass ball, ovulated, and again experienced extended luteal function.This technique offers the advantage of suppressing behavioral estrus because of endogenous progesterone production over an extended period following a single administration of a glass ball. The disadvantage is that it does not work in every mare nor does it appear to have an immediate effect in every mare following intra-uterine placement of the glass ball. However, when it is effective it may serve as an alternative method for suppressing estrous cycle and/or behavior and thus avoiding the need for administration of exogenous progestin products. Readers are cautioned that many questions still exist about the use of intra-uterine glass balls in mares. Further work is required to confirm the efficacy of the use of an intra-uterine glass ball for prolonged luteal function in mares and to identify its mechanism of action.  相似文献