Discussion

There are three major outcomes of these studies. First, treatment of postpartum dairy cows with GnRH and PGF2α in a 7-day period started as early as 21 days effectively induces ovulation. Second, the treatment combination increases the number of luteal phases by day 60 postpartum; and finally, the cyclic ovarian activity that follows the treatment protocol is normal in the majority of the cases.

Following the treatment with GnRH 21 days postpartum in study 1, 87% of the

cows ovulated (in both GP-CL and GP-NCL groups). Similarly, plasma P4 levels ≥ 1ng/ml by 28 days postpartum were observed in 80% of the GP-NCL cows and in all cows of the GP-CL group in study 2. These results are in agreement with the results presented in Chapter 3, and with those reported by others (5, 16, 28).

Two cows in study 1 failed to ovulate after GnRH treatment, while 4 cows in study 2 presumably did so based on the absence of luteal activity by 28 days postpartum.

Both GnRH-anovulatory cows in study 1 were nesting follicles larger than the size at which follicles acquire the capacity to ovulate (10 mm) (data not shown) (83). Failure to ovulate might have been due to a reduced number of LH receptors (45) or caused by a reduced capacity of the follicle to bind to LH due to ongoing atresia (35) as reported previously. The reasons for the ovulatory failure after GnRH treatment in four cows in study 2 are also uncertain. As in study 1, ovulation failure in study 2 might have been due to absence of GnRH-responsive follicles. Three of the four cows that failed to ovulate after the GnRH treatment in study 2 did not have luteal activity for at least 45 days postpartum, and the remaining cow showed its first lutea activity two days after treatment with PGF2α. Nevertheless, the high rate of ovulation after GnRH treatment on day 21 and the proportion of cows with a functional CL by day 28 in both studies further demonstrates, in agreement with the results presented in Chapter 3, that ovulation and synchronization of a new follicular wave can be effectively induced despite differences in the ovarian cyclicity status during this period.

In both studies, the rate of ovulation after PGF2α was lower than the rate of ovulation to GnRH. One reason that limited ovulation after PGF2α was the failure to ovulate in response to GnRH treatment in the three anestrous cows classified as non-synchronized in these studies. In a limited number of cows, the absence of

ovulation by 72 h following treatment with GnRH as observed in the ultrasound-monitored group, and by an increase in P4 levels ≥ 1ng/ml 2 to 5 days after PGF2α treatment in both studies indicated that these cows had ovulations within 4 days prior to PGF2α treatment. Therefore, non-synchronized cows had CL in the very early stages of formation at the time of treatment with PGF2α. Since CL in the early stage of formation are refractory to PGF2α- induced luteolysis (60), the lack of a responsive CL by 28 days postpartum in these cows was the cause of a failed ovulation after PGF2α.

In the GP-NCL (-) cows that ovulated to GnRH treatment in both studies, failure of the CL to regress prevented the occurrence of ovulation after PGF2α treatment.

Interestingly, the P4 levels prior to PGF2α treatment in the GP-NCL (-) group were similar to those in the GP-NCL (+) group, in which luteolysis was successfully induced.

These results indicate that, despite differences in PGF2α-induced regression, the development and the steroidogenic function of the induced CL in the GP-NCL groups were similar. However, the high rate of luteolytic failure of cows in the GP-NCL that had a functional CL by the time of PGF2α treatment under commercial conditions was unexpected. One probable reason for the degree of unresponsiveness between the two studies could have been the heterogeneity of conditions (i.e., animals, environmental) among the four commercial herds. Moreover, pharmacokinetic differences between the two different PGF2α analogues used in either study can not be discarded as a possible cause. The responsiveness of newly formed CL to undergo luteolysis may depend on both the magnitude and the duration of the luteolytic stimuli (70). It is probable that Cloprostenol, based on its long half life (t1/2= 3 hrs) (70), is more resistant to metabolism than Etiproston tromethamine (no published information on t1/2).

Tromethamine salts, which are from natural PGF2α origin have very short half-lives (75).

A shorter milk withdrawal period is recommended for Etiproston tromethamine (12 hours) (Virbac S.A., France) in comparison to Cloprostenol (24 hours) (ASKA Pharmaceutical Co., Ltd. Tokyo, Japan), being the main reason for its use in study 2. In dogs, Etiproston tromethamine has been reported to be less effective than Cloprostenol for CL regression when administered in a single dose (50).

Interestingly, luteolysis was unanimously synchronized in the GP-CL (+) groups in both studies. This result is in accordance with reports showing improvement in the response to hormonal synchronization programs in cows maintaining luteal activity (62, 63). Probably, the reason for the presence of CL prior to PGF2α and the uniform luteolytic response in this group was because cows were in the early luteal phase when the treatment protocol started. The optimum suggested stage for the start of a GnRH- PGF2α synchronization protocol in cycling cows is the period between days 5 and 10 of the estrous cycle (62)

The treatment with the GnRH- PGF2α protocol to early postpartum dairy cows reduced the time postpartum to recovery of cyclic ovarian activity in both GP-CL (+) and GP-NCL (+) groups. Moreover, the mean interval in days from the PGF2α-induced luteolysis to the beginning of the two consecutive luteal phases (2nd and 3rd P4 rises, respectively) observed under research conditions was precisely replicated in the field.

These results confirm that ovulation after GnRH and the synchronization and ovulation of a new follicle after the PGF2α treatment did occur in the GP-CL (+) and the GP-NCL (+) groups in the commercial dairy herds. Since days 5-9 of the estrous cycle yield the best synchronization and fertility results after timed appointed breeding (70, 97), the synchronous start of a luteal phase close to the end of the voluntary waiting period as herein reported may be a prospective base for non-estrus detection based breeding

strategies.

The length of the estrous cycle following the protocol in cows that responded with ovulation after both treatments was normal in the majority of the cases. Moreover, estrous cycles of long durations were significantly reduced to the point of being almost absent in treated cows. In contrast to a previous report in which normal cycles were reduced when the treatment with GnRH preceded PGF2α 10 days (5), our present findings may suggest a 7 day period as optimum for the synchronization of ovulations and the cyclic ovarian activity when similar pharmacological treatments are to be used during this period.

Based on the number of exposures to P4 within 60 days postpartum, a significant acceleration of the cyclic ovarian activity occurred under field conditions in the C-CL group (2.7 luteal phases) as well as in the GP-CL (+) and GP-NCL (+) groups (2.6 and 2.8 luteal phases, respectively). A similar tendency was observed under research conditions suggesting that, in this case, the small number of animals was a limiting factor to obtain comparative results among the studies. Furthermore, in these treated cows, hormonal treatment induced P4 profiles that mimicked the ovarian activity of the C-CL group which had an early spontaneous first ovulation.

Similarly and in agreement with previous reports, the length of ovarian cycles were normal in cows that responded with an early first ovulation after GnRH treatment only (PGF2α-refractory cows in our study) (5). Others have previously reported an increase in uterine infections and pre-breeding anestrous when cows are treated only with GnRH in the early stages of postpartum (20). Although, the length of the ovarian cycles in both studies could be confirmed in 11 cows responding to GnRH only, cyclicity after the treatment period was normal in almost all cows. Most probably, the

uterine involution of cows in these experiments was either normal or was improved by the GnRH treatment as reported previously (9, 22). Together, these results accounted for the overall greater proportion of normal cycles in treated cows.

Evaluation of the subsequent fertility considering the cyclicity status 21 and 28 days postpartum revealed that the presence of active CL 28 days postpartum and CL regression and ovulation following the treatment with PGF2α are critical factors that increase fertility. An overall occurrence of significant improvement in the parameters of fertility in the treated cows seemed to have been compromised by CL refractoriness to the PGF2α treatment. The use of prostaglandins with a longer biological half life and/or the administration of two consecutive doses may help to correct this problem. Two treatments with PGF2α at 8-h intervals improved fertility outcomes in dairy cows (75).

In conclusion, the treatment of dairy cows with a 7-day GnRH-PGF2α based protocol as early as 21 days postpartum induces ovulation and is capable of synchronizing the early return to cyclic ovarian activity in both early ovulating cows and cows having the risk for a prolonged postpartum anestrous period. The presence of active CL 28 days postpartum and CL regression and ovulation following the treatment protocol are critical factors that increase fertility.