SYNCHRONIZED OVULATION WITH TIMED INSEMINATION VERSUS EXOGENOUS PROGESTERONE WITH INSEMINATION AT AN INDUCEDESTRUS AS THERAPEUTIC STRA TEGIES FOR OVARIAN CYSTS IN LACTATING DAIRY COWS By MARY BRONWYN CRANE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005
ii ACKNOWLEDGMENTS I wish to express my sincere appreciation to: Dr. Louis F Archbald (my advisor), Prof essor of Veterinary Medicine, for his support and encouragement, the passing of know ledge and skills, for the opportunity to teach and to learn from his example, for his enth usiasm and his genuine care and concern. Dr. Carlos A Risco, Professo r of Veterinary Medicine, for his valuable advice and assistance with the project, his encouragement and enthusiasm. Dr. Pedro Melendez, Assistant Professo r of Veterinary Medicine, for his willingness to help at all times, his enthusia sm, and his valuable assistance with the statistics and the project. Dr. Albert de Vries, Assistant Professor of Animal Science, for his teaching and work in developing an economical model fo r this data, his valuable comments and support. Mr. Ingo Kreig, owner of Mecklenburg Dair y, for his cooperation with the project, particularly for the use of his cows and facilities. Mr. Roger Rowe, farm manager at Meckle nburg Dairy, for his time and assistance with the project. Florida Dairy Check-Off program for the economic support of the project. Dr. Julian Bartolome, for his valuable co mments and assistance with the project. Dr. Mauricio Benzaquen and Dr. Katherin e Hendricks, fellow graduate students, for their time and assistance with the project.
iii Mr. James Burrow, for his assistance with the progesterone assays. Faculty, residents and staff of the Food Animal Reproduction and Medicine Service for their interest, teaching and assistance. My parents, brother and sisters fo r their support a nd encouragement. My husband, Paul, and my son, Jude, for their love and support over the years.
iv TABLE OF CONTENTS page ACKNOWLEDGMENTS..................................................................................................ii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES..........................................................................................................vii ABSTRACT.....................................................................................................................vi ii INTRODUCTION...............................................................................................................1 LITERATURE REVIEW....................................................................................................4 Gonadotropin-Releasing Hormone...............................................................................4 Introduction...........................................................................................................4 Chemical Nature of the Hormone..........................................................................4 Control of GnRH Secretion...................................................................................5 Function of GnRH...............................................................................................14 Activity on gonadotropes.............................................................................14 Extrapituitary functions of GnRH................................................................19 Clinical Applications...........................................................................................20 Folliculogenesis..........................................................................................................22 Cystic Ovarian Disease in the Dairy Cow..................................................................31 Introduction.........................................................................................................31 Definition and Diagnosis.....................................................................................32 Incidence Rate and Risk Factors.........................................................................35 Etiology...............................................................................................................37 Fate and Endocrinology.......................................................................................40 Pathogenesis........................................................................................................42 Treatment.............................................................................................................51 Economics of Cystic Ovarian Disease................................................................59 Timed Insemination Programs in the Dairy Cow.......................................................63 Programs using Prostaglandin F2 Alpha.............................................................64 Programs using Prostaglandin F2 Al pha and GonadotropinReleasing Hormone..........................................................................................................68 Programs using Progesterone..............................................................................74 MATERIALS AND METHODS.......................................................................................80
v RESULTS........................................................................................................................ ..85 Baseline Data Comparisons........................................................................................85 Likelihood to be Inseminated, Return to Cyclicity, Conception and Pregnancy Rates.......................................................................................................................87 Likelihood to be Inseminated..............................................................................88 Return to Cyclicity..............................................................................................89 Conception and Pregnancy Rates........................................................................94 DISCUSSION....................................................................................................................9 8 LIST OF REFERENCES.................................................................................................116 BIOGRAPHICAL SKETCH...........................................................................................140
vi LIST OF TABLES Table page 1. Reported incidence rates for cystic ovarian diseas e in the literature.............................36 2. Reported stages of lactation with an incr eased incidence of cystic ovarian disease.....36 3. Baseline data comparisons for cows in the Ovsynch and CIDR groups........................85 4. Summary of heat detection rate, conception rate, pregna ncy rate and missing values/ lost cows by group............................................................................................................. 87 5. The risk of insemination for cows in the Ovsynch and CIDR groups adjusted for parity, DIM, BCS and milk produ ction on the day of diagnosis.......................................89 6. The risk for the presence of a CL on Day 21 for cows in the Ovsynch and CIDR groups adjusted for parity, DIM, BCS and milk production on the day of diagnosis.......90 7. Descriptive statistics for progesterone values (ng/ml) based on the presence or absence of a CL................................................................................................................ ..91 8. Contingency table for the di agnosis of a CL based on pa lpation per rectum and U/S (RP+U/S) or progesterone value gr eater than or equal to 1.0ng/ml..................................92 9. Progesterone level as it relates to pregna ncy in cows diagnosed with a CL on day 21............................................................................................................................. ...........92 10. Least squares means and standard error of the mean for progesterone concentrations and associations with the explanatory variables........................................93 11. The risk of conception for cows in the Ovsynch and CIDR groups adjusted for parity, DIM, BCS and milk produ ction on the day of diagnosis.......................................95 12. The risk factors associated with pregnancy rate for co ws in the Ovsynch and CIDR groups adjusted for parity, DIM, BCS and milk production on the day of diagnosis.......97
vii LIST OF FIGURES Figure page 1. The distribution of milk production on the day of diagnosis (Kg) for cows in both groups combined................................................................................................................ 86 2. The proportion of cows per group that were inseminated (AI), returning to cyclicity, conception rate (CR), and pre gnancy rate (PR) for the Ov synch group (white bars) and the CIDR group (black bars)..............................................................................................87 3. The proportion of cows inseminated (whi te) compared to cows not inseminated (black) for the Ovsynch and CIDR groups........................................................................89 4. Box and whisker plots of progesterone valu es for cows with (CL) or cows without (No CL) a CL................................................................................................................... ..91 5. Distribution of progesterone concentrations (labeled P4 value; ng/ml) on Day 21 for cows in the Ovsynch (labeled Gr 1) and CIDR (labeled Gr 2) groups...................94 6. Conception rate (CR) and pregnancy rate (PR) divided by primiparous (1st Lactation; white) and multiparous (2nd + Lactation; black) for cows in the Ovsynch and CIDR groups...............................................................................................................9 6 7. Conception rate by milk production categ ory for cows in the Ovsynch and CIDR group.......................................................................................................................... ........96
viii Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science SYNCHRONIZED OVULATION WITH TIMED INSEMINATION VERSUS EXOGENOUS PROGESTERONE WITH INSEMINATION AT AN INDUCED-ESTRUS AS THERAPEUTIC STRA TEGIES FOR OVARIAN CYSTS IN LACTATING DAIRY COWS By Mary Bronwyn Crane August 2005 Chair: Louis F. Archbald Major Department: Veterinary Medical Sciences Therapeutic strategies for bovine ovarian cysts could involve either the use of GnRH or exogenous progesterone; however th ere is no information available concerning the comparative efficacy of these two treatm ent strategies. The purpose of this study was to compare the clinical effectiveness of the Ovsynch and CIDR protocols under commercial conditions for the treatment of cystic ovarian disease in dairy cattle. A total of 401 lactating dairy cows with ovarian cysts were enrolled in the study from October 13, 2003, to September 20, 2004. A ll cows diagnosed with ovarian cysts were alternatively allocated to 2 treatment groups on the day of diagnosis. Cows in the Ovsynch group were treated with GnRH on Day 0, PGF2 on Day 7, GnRH on Day 9, and timed inseminated 16-20 h later. Cows in the CIDR group were treated with a CIDR Insert on Day 0 for 7 days. On Day 7, the CIDR was removed, and cows were treated with PGF2 All cows in the CIDR group were obs erved for estrus and cows exhibiting
ix estrus within 7 days following removal of the CIDR and PGF2 administration were inseminated. The outcomes of interest for this experiment were the likelihood to be inseminated, the presence of a CL on Day 21, conception rate and pregnancy rate. Data for these variables were analyzed using logist ic regression adjusting for parity, days in milk, body condition score, season, and milk production on the day of diagnosis. The percentage of cows inseminated in the Ovsynch and CIDR groups were 82% and 44%, respectively. The odds for cows in the Ovsynch group to be inseminated were 5.6 times more than the odds for cows in the CIDR group. The percentage of cows with a CL on Day 21 for the Ovsynch and CIDR groups was 83% and 79%, respectively. The odds for cows in the Ovsynch group to have a CL on Day 21 were 2.2 times more than the odds for cows in the CIDR group. Despite the decreased likelihood for a CL, cows in the CIDR group had higher progesterone c oncentrations on Day 21. The conception and pregnancy rates for cows in the Ovs ynch group were 18% and 14%, respectively. Conception and pregnancy rates for cows in the CIDR group were 23% and 9.5%, respectively. There was no signi ficant effect of treatment on conception or pregnancy rate. The odds for primiparous cows to conceive were 4.1 times more than the odds for multiparous cows. The odds for cows in the 3rd and 4th quartiles of milk production on the day of diagnosis to conceive were more than the odds for cows in the lowest quartile of milk production. The results of this study indi cate that fertility is not di fferent between cystic cows treated with either the Ovsynch or the CIDR protocol, although cows treated with the Ovsynch protocol were more likely to return to normal cyclicity.
1 INTRODUCTION Bovine ovarian cysts are follicles that fail to ovulate at the time of estrus (Garverick, 1999). This condition presents an enormous economic problem because these cows are infertile as long as the condi tion persists (Kesler a nd Garverick, 1982). It has been reported that the interval from part urition to conception in cows with ovarian cysts is 64 days longer than that observed in cows without ovarian cysts (Bosberry and Dobson, 1989). Since the cost of each day a cow is not pregnant beyond the voluntary postpartum waiting period (usually 70 days post partum) is approximately $2.50, a cow with ovarian cysts will incur an additional cost of $55 to $160 per lactation (Bartlett et al., 1986). Other costs of this condition incl ude an increased culling rate, semen, and veterinary costs. This has been calculated to be approximately $75 per case. Therefore, the total cost of this cond ition has been estimated to be $130 to $235 per cystic cow per lactation (Bartlett et al., 1986) The economic magnitude of this condition becomes more obvious in large (1,100-3,100 milking cows) da iry herds with an incidence of 9-25%. The exact cause of ovarian cysts is no t presently known, but it appears that an important component in the pathogenesis of this condition is the inappr opriate, or lack of, release of gonadotropin-releasing hormone (GnR H) at the time of estrus. It has been suggested that an underlying mechanism in th e development of ovarian cysts involves a hypothalamic lesion which causes follicular estr ogen to be ineffective in inducing a GnRH/LH surge at the time of estrus (G umen and Wiltbank, 2002), and that this hypothalamic lesion could invol ve the estrogen receptor (ER ). Further, it has been
2 speculated that treatment with progesterone may induce the ER in the mediobasal hypothalamus which will foster a GnRH/LH su rge in response to follicular estrogen (Gumen and Wiltbank, 2002). Collectively, this information suggests that therapeutic strategies for bovine ovarian cysts could involve either the use of GnRH or exogenous progesterone. In fact, several experimental protocols using these hormones have been shown to be effective in the treatment of bovine ovarian cysts. A protoc ol involving multiple injections of GnRH, prostaglandin F2 (PGF2 ), and timed insemination without detection of estrus has been shown to be effective in treating this condi tion (Bartolome et al., 2000). However, this protocol is labor-intensive and time-consuming since it involves handling of cows multiple times and a 10-day period of time. In the USA, use of exogenous progesterone for treating bovine ovarian cysts has been ha mpered by lack of FDA approval for the use of progesterone in lactating dairy cows. However, an intravaginal device containing progester one (EAZI-BREED CIDR) was recently approved by the FDA for us e in lactating dairy cows. This protocol is relatively simple, does not involve as much handling of cows, and is not as laborintensive and time-consuming as the one using GnRH and PGF2 Therefore, it could be a more acceptable clinical approach for the treatme nt of ovarian cysts in the lactating dairy cow. However, there is no information avai lable concerning the comparative efficacy of these two treatment strategies in a single, large dairy herd. The conceptual hypothesis of this study was that lactating dairy cows with ovarian cysts treated with exogenous progesterone and a luteolytic dosage of PGF2 which are inseminated at an induced-estrus will have a different pregnancy rate compared to cows
3 with ovarian cysts which are subjected to synchronization of ovulation and timed insemination without de tection of estrus. The objective of this study was to compare the effectiveness of these two protocols in a single, large (1,500 milking cows) commer cial dairy herd. By using a single, large herd dairy herd, we can obtain the large num ber of cows needed to show potential statistically significant differences, and be assu red that all cows will be subjected to the same managerial and nutritional conditions during the period of the study. In this approach, the only experimental variable will be the treat ment protocols.
4 LITERATURE REVIEW Gonadotropin-Releasing Hormone Introduction Gonadotropin-releasing hormone (GnRH, also called luteinizing hormonereleasing hormone or LHRH) is a pivotal reproductive hormone in both the male and female. It is produced in the hypothalamus a nd released from the median eminence under the control of various neuronal inputs and gonada l steroids. Once rele ased at the median eminence into the hypophyseal portal vessels, it traverses to the anterior pituitary gland where it binds to receptors on gonadotr opes (Ojeda and McCann, 2000). Once bound to the gonadotrope, it stimulates the release of LH and FSH. These two glycoportein hormones control folliculogenesis, ovulati on, CL function and steroidogenesis. These functions provide steroidal f eedback to the hypothalamus which ensures the continuation of normal ovarian cyclicity. Chemical Nature of the Hormone GnRH is a decapeptide consisting of pG lu-His-Trp-Ser-Try-G ly-Leu-Arg-Pro-Gly, listed from the carboxyl to amino terminal. Th e highly conserved aspe cts of the peptide throughout evolution include the length (10 amino acids), the 1st four amino acids at the carboxyl-terminus, and the final two amino acids at the amino-terminus (Millar et al., 2004). The high degree of conser vation indicates the importan ce of these characteristics for receptor binding and activation. The amino-te rminus is essential for receptor binding and activation while the carboxyl-terminus is only essential for receptor binding.
5 Substitution of constituent amino acids at the am ino-terminus can lead to a hormone with antagonistic properties (M illar et al., 2004). In humans, GnRH originates from a 92 amino acid-precursor peptide and after proteolytic cleavage the decapeptide (GnRH) is associated with a 56 amino acid-peptide (Ojeda and McCann, 2000). A second variant of GnRH with unknown function has been found and differs by 3 amino acids. This new GnRH variant is called GnRH II and is mostly located in peripheral tissues such as the kidney, bone marrow and prostate (Millar et al., 2004). GnRH is packaged in storage granules which are transported down axons to the median eminence where their release is coordinated in a pulsatile manner. Control of GnRH Secretion GnRH is released in a pulsatile manner from the median eminence at low and infrequent levels throughout th e luteal phase of the estrous cycle. During proestrus, the preovulatory estradiol surge causes an increase in the frequency of GnRH pulses, which result in an LH surge from the anterior pitu itary gland. There is ge nerally considered to be a negative correlation between GnRH/LH levels and progesterone concentration. This differential pattern of GnRH release during various stages of the estrous cycle is controlled by distinct GnRH neuronal populatio ns and the different influences of gonadal steroids. In order to characterize the involvement of ovarian steroids in the regulation of LH secretion, Karsch et al. (1980) performed a se ries of experiments in ewes which were intact, ovariectomized (OVX), OVX and pr ogesterone supplemented, OVX and estrogen supplemented, or OVX and estrogen plus progesterone supplemented. Luteinizing hormone concentrations were monitored duri ng the study period of one estrous cycle. A significant finding in that study was that ovariectomized ewes had elevated LH
6 concentrations which could be partially redu ced by either estrogen or progesterone alone, but required the combination of both steroids to reduce LH to similar concentrations observed in intact animals. Although both gonadal steroids have an inhi bitory effect on LH (Karsch et al. 1980), their influence over many days on the pulsatile character of LH is another important component of normal ovarian cyc licity. During the luteal phase, LH pulses occurred approximately every 3.5 hours and each pulse was followed by an increase in estradiol (Baird, 1978). Similarly, OVX ewes tr eated with progesterone for 10 days had decreased LH pulse frequency and increased amplitude compared with OVX controls (Goodman and Karsch, 1980). After removal of progesterone there was an increase in both LH and estradiol concentrations, followed shortly by estrus (Kar sch et al., 1979). In cyclic ewes, follicular phase LH pulses have an incr eased frequency and decreased amplitude, which results in an overall increase in basal LH concentr ation (Baird, 1978). This was verified experimentally when long-term subcutaneous estradiol implants in the absence of progesterone caused a decrease in LH pulse amplitude yet maintained a frequency similar to OVX controls (Karsch et al., 1980). This expe riment also verified that the withdrawal of progesterone alone was not sufficient to induce the LH surge and estradiol alone was not sufficient for the full magnitude preovul atory LH surge. Despite the decreasing amplitude of LH pulses during the follicular phase, the estradiol pr oduced by the follicle was increased in comparison to responses obs erved during the luteal phase (Baird, 1978). This suggests that, during the follicular phase, the follicle has an increased sensitivity to LH.
7 There is evidence that LH is responsible for the rising estradio l concentrations and the preovulatory increase in both of these hormones can be prevented by the administration of progesterone (K arsch et al., 1979). Similar results were also found in proestrus rats through measuring the concen tration of LH releas ing-hormone (LHRH) instead of LH. Animals which were ovariec tomized and treated with estradiol had a similar LHRH surge to intact proestrus rats while the two groups lacking estradiol (OVX controls and OVX progesterone treated rats ) had significantly d ecreased LHRH surge (Sarkar and Fink, 1979). To further validate th e involvement of estradiol in the GnRH surge, the expression of c-fos, a protooncoge ne used as a marker of cellular activation, was examined in GnRH cells during the diffe rent phases of an induced estrous cycle (Moenter et al., 1993). The resu lts indicated that, while th ere was little c-fos activity during most of the estrous cycle, estradio l induced c-fos expression in 48% of GnRH neurons and many other non-GnRH hypothalami c neurons, indicating estradiol induced cellular activation. Collectively, these studies suggest that LH is involved in increasing estradiol concentrations and estradiol feeds forward to in crease LH pulse frequency and ultimately triggers the GnRH/ LH surge. Further characterization of the GnRH su rge indicated that GnRH pulse frequency increased while amplitude decreased during the preovulatory phase and this cumulates into a GnRH surge with concentrations 40 times greater than basal (M oenter et al., 1991). The GnRH surge is characterized by an increa se in pulse frequency which causes an LH surge and finally outlasts the LH surge (Moe nter et al., 1991). During the GnRH surge there is an increase in the GnRH interpulse concentrations, which is built upon by the increased pulse frequency, until finally GnRH release switches from a pulsatile pattern
8 into a continuous pattern (Car aty et al., 1995). The entire GnRH surge is required for an LH surge of normal size and duration and if the surge is prevented during the ascending or descending phases, the LH surge will also abruptly stop (Evans et al., 1996). Furthermore, once the surge has been trigge red, the ovaries (or gonadal steroids) are not required for further hypothalamic stimulation to main tain the surge (Webb et al., 1981). Many studies have focused on mapp ing the neuronal network involved in regulating GnRH secretion. This has involved examining the location and distribution of GnRH neurons, the distribution of neurons c ontaining estrogen receptors (ERs) and the interneurons and neurotransmittors involved in relaying the signal between the estrogen responsive and GnRH containing cells. Other st udies have focused on the variation in the activity of these neurons during different phase s of the estrous cycle. In the ewe, GnRH neurons projecting to the median eminen ce were traced using florescent tracers and located to the diagonal band of the Broca/medi al septal region, me dial preoptic area, anterior hypothalamic area and the medial basal hypothalamus (Jansen et al., 1997). The percentages of the total GnRH population w ithin each area did not differ significantly (Jansen et al., 1997). In an earlier study preformed during the luteal phase, LHRH neurons were found throughout the medial basa l hypothalamus with 5% interacting with other LHRH neurons (Leshin et al., 1988). In the majority of studies, GnRH neur ons have been found to lack estrogen receptors (Shivers et al., 1983; Lehman and Karsch, 1993). Then using reverse transcriptasepolymerase chain reaction (R T-PCR) technology in immortalized mouse GnRH neurons, Shen et al. (1998) found ev idence for biologically active estrogen receptors. In the ewe, the ventromedial nucleus has estrogen receptors which are up-
9 regulated by progesterone and involved in tran smission of the GnRH surge (Blache et al., 1994). These neurons, which are involved in estrus and the GnRH surge, secrete somatostatin (Herbison, 1995). Neurons contai ning estrogen receptors in different areas secrete different neurotransmittors. Specifically 30% of ER neurons in the preoptic area secrete the inhibitory neurotransmitter -aminobutyric acid (GABA; Herbison, 1995), while in the arcuate nucleus only 3-5% secreted dopamine and 15-20% secrete endorphin ( -END; Lehman and Karsch, 1993). Alt hough only a small percentage of ER neurons exhibited co-expression of a neurotransmitter, many of them were in close contact with ER containing cells (Lehman and Karsch, 1993). The neuronal path between th e input of steroidal horm ones and GnRH release is complicated and involves th e interplay of many exci tatory and inhibitory neurotransmitters. This area of research is rapidly evolving, while development of a reliable model is compounded by species differe nces in the control of GnRH release. The neurotransmitters generally considered excitatory on GnRH neurons (stimulate production and release of GnRH) include neur otensin (NT), norepinephrine, leptin and galanin; while neuropeptide Y (NPY), dopamine, GABA, and orexins can have a stimulatory or inhibitory effect depending on the steroidal milieu. Neurotransmitters with a consistent inhibitory effect include somatostatin, interleukin-1 neuropeptide K, and the endogenous opioid peptides: -END and dynorphin. Neurotensin is a neuropeptide possibly i nvolved in transmission of the estradiol induced GnRH signal in rats. Smith and Wise (2001) found that there was an increase in NT biosynthesis on the morning of proestru s which was necessary for the preovulatory surge. In the same study, eighty perc ent of GnRH neurons expressed NT
10 immunoreactivity. The increase in NT bios ynthesis observed during proestrus was paralleled by the estradiol surge, suggesting it may play a role in increasing NT expression (Smith and Wise, 2001). The endogenous opioid peptides, such as -END and dynorphin, are involved in suppression of LH. Although -END has been colocalized with ERs (Lehman and Karsch, 1993), the role of steroi ds in mediating the release of -END during the estrous cycle is not clearly define d. Neurons producing preopiomel anocortin, a precursor to END, were found interacting with 6% of LH RH neurons (Leshin et al., 1988). There was also extensive intermingling of these two neur on types at the zona externa of the median eminence, suggesting that -END acts at both the GnRH cell bodies and axon terminals in mediating the release of GnRH (Leshin et al., 1988). This is supported by findings of increasing -END concentrations in the median em inence from the luteal to follicular phase of the estrous cycle, then finally declining -END concentrations during the LH surge (Herbison, 1995). It has been suggested that -END may play a role in preventing the premature activation of GnRH neurons during the follicular phase (Herbison, 1995). Opioid antagonists have been shown to s timulate LH secretion (Herbison, 1995). Recent data shows that estrogen and proge sterone are likely both involved in the regulation of -END activity at -receptors in the preopt ic area and medial basal hypothalamus for the generation of the GnRH surge (Goodman et al., 2004). Furthermore, there is evidence to suggest that the role of progeste rone in regulating LH pulse frequency is mediated through dynorphin activity at -receptors in the medial basal hypothalamus (Goodman et al., 2004). At these receptors, progesterone induced dynorphin acts to slow LH pulse frequency.
11 The inhibitory neurotransmitter, GABA, is released in synchrony with GnRH pulses indicating an involvement with LH pulsatility (Kalra et al., 1997). Although GABA is usually an inhibitory neurotransmitter, it can be stimulatory on adult GnRH neurons through the GABAA receptor (Sullivan and Moenter, 2004). Progesterone alters GABA secretion in the preoptic area (Scott et al., 2000) and a fall in GABA concentration is associated with the LH surge (Herbison, 1995). It was recently demonstrated that progesterone reduced bot h the frequency and size of GABAA receptor mediated presynaptic currents in mice, and th is caused a reduction in pulsatile GnRH release (Sullivan and Moente r, 2005). In the same study, dihydrotestosterone had the opposite effect and increased GAB Aeric tone. It also caused an increase in the number of synaptic contacts between GABAergic and Gn RH cells. An interaction between the metabolic hormones leptin, neuropeptid e Y, and endogenous opioids with GABA indicates that GABA is involved in relaying information regard ing metabolic status to the reproductive axis. Leptin increases GABA ergic tone to GnRH neurons, providing stimulation for GnRH release in both fed and fasted animals, while NPY and endogenous opioids decreased GABAergic tone in fed animals only (Sullivan and Moenter, 2004). Metabolic hormones with effects on the reproductive axis incl ude insulin, leptin, orexins A and B, and NPY. Insulin is releas ed in response to high blood glucose and acts to up-regulate cellular glucose intake but it is also a signal of satiet y. Leptin is a satiety factor released by adipocytes. It increases w ith feeding and fat content and is generally considered to have a positive effect on re production (Houseknecht et al., 1998). Orexins A and B are oerexogenic factors released fr om the lateral hypothalamic area and are involved in hunger regulation. Ne uropeptide Y is an orexogeni c neuropeptide released
12 during fasting and is generally considered a strong inhibitor of GnRH and LH release (Gazal et al., 1998). However, other eviden ce suggests that NPY may be involved in stimulating the GnRH surge in an estr ogen dependent manner (Xu et al., 2000). Neuropeptide Y can have a positive effect on Gn RH secretion in the presence of steroids by increasing galanin concentration, but in the absence of steroi ds, it can increase -END concentration and have a negative effect on GnRH release (Kalra et al., 1997). Galanin and NPY can also act at the level of the ante rior pituitary gland to increase gonadotrope response to GnRH (Kalra et al ., 1997). In contrast to these stimulatory effects, a dose of 500ug of NPY in cows caused immediate ce ssation of LH release for 4 hours and disruption of GnRH for 1.5 to 3 hours (Gazal et al, 1998). The role of leptin in peripubertal anim als was examined in a study using normal and nutritionally restricted heif ers (Zieba et al., 2004). In these heifers, leptin caused a slight increase in mean LH concentrations and did not affect pulse frequency. The effects of insulin and leptin on GnRH secretion have been examined in vitro using cultured rat hypothalamic sections (Burcelin et al., 2003). In that study, addition of insulin resulted in a dose-dependent increase in GnRH secreti on, with the first detectable significant difference occurring at insulin concentrati on 139% of baseline. In glucose controlled mice, hyperinsulinemia caused a 50-60% increase in LH concentrations (Burcelin et al., 2003). When leptin was added to the hypotha lamic culture, there was no detectable increase in GnRH, but leptin did cause a si gnificant increase in the GnRH response to insulin, suggesting a potentiation effect (Burcelin et al., 2003). In nutritionally stressed cows, leptin caused an increase in the mean size of GnRH pulses and increased LH pulse amplitude (Zieba et al., 2004). These results suggest that leptin can act at both the
13 hypothalamus to increase GnRH release and at the anterior pituitary to sensitize gonadotropin response to GnRH. It has also b een suggested that fa sting sensitizes the reproductive axis to the effects of leptin a nd this may occur through an up-regulation of leptin receptors in the hypot halamus (Zeiba et al., 2004). Orexin-containing fibers have been co-l ocalized with 75-85 % of GnRH neurons (Campbell et al., 2003). Of these GnRH neur ons co-localized with orexin containing fibers, 85% contained the orex in A specific receptor, OX-R 1, which indicates a possible functional relationship. The OX-R1 recepto r found on GnRH cells is generally a stimulatory receptor type suggesting that or exins may have a direct positive effect on GnRH secretion (Campbell et al., 2003). The effect of steroids on the orexin-induced stimulation or inhibition on Gn RH cells has not been fully de termined. It is possible that orexins may inhibit GnRH release through i ndirect pathways, while stimulating GnRH through a direct pathway, depending of the steroi dal milieu (Campbell et al., 2003). Another complicated explanation is the i nvolvement of progesterone on both the tonic GnRH pulses and the in itiation of the GnRH surge. MacLusky and McEwen (1978) identified two different populations of proge sterone receptors (PR): 1) receptors in the cerebral cortex and midbrain, and 2) recepto rs in the hypothalamus, pituitary, preoptic area, and uterus. The latter group of progest erone receptors was induced and up-regulated by estrogen (MacLusky and McEwen, 1978). Furt hermore, the progesterone receptors that were up-regulated by estrogen prior to the preovulatory LH surge seemed to be induced in the absence of ligand (Levine, 1997). Estrogen also e nhances progesterone inhibition on tonic LH secre tion by up-regulating PRs in th e ventromedial nucleus and arcuate nucleus (Scott et al., 2000). Ther e is evidence that th e estradiol-induced
14 progesterone receptors and NPY work togeth er to produce the GnRH surge (Xu et al., 2000). The NPY receptor, Y1r, stimulates GnRH release when bound to ligand. The estrogen-induced PRs in the arcuate nucleus ca n be activated either by progesterone or trans-activated by other neurotransmitters. On ce activated these PRs act to increase the amount of Y1r mRNA and therefore amplify th e signal created by NPY (Xu et al., 2000). Function of GnRH Activity on gonadotropes Once GnRH is released at the median emin ence, it enters the portal vessels and is carried to gonadotropes in the anterior p ituitary gland. The gonadot ropes synthesize and release LH and FSH in response to GnRH bi nding at its membrane receptor. The relative response of the gonadotropes to GnRH can be affected by many different hormones including estradiol, progesterone, inhibin, activin and follistatin (Caraty et al., 1995). The GnRH receptor is a 7 transmembrane, G-protein coupled receptor. Once activated by binding with GnRH, guanine di phosphate (GDP) is exchanged for guanine tripohosphate (GTP) on the associated G-prot ein complex (Anderson, 1996). This causes the G subunit to dissociate from the G subunits and activate the intercellular messenger systems involved in the synthesis and release of LH and FSH. Although there is evidence for multiple intracellular messe ngers after receptor act ivation, the primary pathway is through activation of phospholip ase C (PLC; Anderson, 1996). Activation of PLC increases intracellular i nositol 1,4,5 triphosphate (IP3 ) and diacylglycerol (DAG), IP3 then causes an increase in intracellula r calcium and DAG activates protein kinase C (Anderson, 1996). Together these factors acti vate the calcium dependent exocytotic mechanisms leading to the release of LH a nd FSH, which are packaged into distinct secretory granules.
15 The magnitude of the FSH release is less th an that of LH and it appears that FSH release is constitutive while LH release is more highly regulated by GnRH (Anderson 1996). Estradiol and follistatin suppress FSH and are likely involved in its tonic secretion (Moss et al., 1981). In OVX and luteal phase ewes, LH and GnRH pulses were highly correlated while FSH pulses matched with only 50% and 25% of GnRH pulses (Padmanabhan et al., 2003). The proportion of FSH and GnRH pulses that were in concordance was considered a statistically ra ndom association in that study. Furthermore, the pulsatility of FSH was not affected by administration of a GnRH antagonist. These results suggest there is a com ponent of the episodic FSH secr etion that is independent of GnRH (Padmanabhan et al., 2003). These author s speculated that th e continued episodic release of FSH may be due to any one of the following possib ilities; incomplete receptor blockade, an intrinsic pituitary FSH rhythm icity, a 2nd class of GnRH receptors, the paracrine activity of activins, inhibins and follistatin, or an unknown hypothalamic FSH releasing factor. The effect of estradiol on pituitary res ponsiveness to GnRH has been examined in many species. In a study using anestrous ewes pretreated with a bolus of differing doses of estradiol, Reeves et al. (1971) found that the higher dose (250ug) of estradiol increased pituitary responsiveness to exogenous LHRH. De spite the potential eff ect at the pituitary, these authors could not rule out that estradiol was also acting at the hypothalamus to increase endogenous GnRH. Adams et al., ( 1975) found that estrogen pretreatment during Days 10 and 11 of the estrous cycle followe d by pulses of LHRH administered every 2 hours for 72 hours, caused an overall decr eased pituitary res ponse to the LHRH compared to oil treated controls. This was characterized by an initial increase in LH
16 which then diminished to near zero by one to two days. Later, using OVX ewes pretreated for 10 days with progesterone, es tradiol or an empty implant, Goodman and Karsch (1980) found that ewes pretreated with estradiol had a slightly prolonged but significantly decreased LH surge in response to exogenous GnRH compared to the other two groups. These results lead to the question of whether or not prolonged administration of estradiol can be inhibito ry at the pituitary. A recent study using OVX rats found that prolonged estradiol exposure caused an abolition of LH surges, but that this was due to a reduction in the percentage of activated GnRH neurons in the hypothalamus (Tsai and Legan, 2002). The postulated effects of steroids on th e GnRH induced rel ease of LH and FSH included alterations in pituitary stores of these hormones and regulation of GnRH receptors (GnRH-R). The literatu re is not in complete agreement with respect to these mechanisms. One study found that pretreatme nt with progesterone had no effect on pituitary LH/FSH stores, but inhibited LH by decreasing hypothalamic release of GnRH (Moss et al., 1981). The same authors also f ound that estradiol decreased pituitary stores of LH and FSH, decreased FSH release and increased GnRH-R. Another study using hypothalamic-pituitary disconnected ewes pretre ated with estradiol found the subsequent administration of progesterone directly i nhibited pituitary LH response to exogenous GnRH (Grimus and Wise, 1992). This work sugges ts that estradiol sens itizes the pituitary to progesterone negative feedback, po ssibly through the estr adiol dependent upregulation of progesterone receptors. Since gonadotrope response to GnRH is partially dependent on the number of GnRH receptors, numerous studies have work ed to characterize changes in receptor
17 numbers as a result of gona dal steroid input. As previ ously mentioned, Moss et al., (1981) found that estradiol increased pituitar y GnRH-R content. Ka iser et al. (1993) examined the effect of estradiol and testoste rone replacement therapy in gonadectomized rats. They found that both steroids caused a reduction in the post -gonadectomy rise in GnRH-R, with estradiol resulting in a relativ ely greater reduction in receptor mRNA than testosterone. The role of proge sterone in the regulation of GnRH-R was examined in the ewe after luteolysis. After th e reduction in peripheral proge sterone following luteolysis, there was an increase in GnRH-R mRNA by 12 hours and an increase in GnHR-R by 24 hours (Turzillo et al., 1994). During this study period estradiol and ER levels remained constant, suggesting the change in GnRH-R was due to decreased progesterone, or the removal of a negative feedback influence. These authors could not exclude the possibility that increased GnRH pulses may also be involved in up-regulating its own receptor. The differential effects of pulsatile versus continuous GnRH administration on gonadotopin secretion is anothe r area under investigation. Many studies have observed the phenomenon of declining LH release after frequent or continuous GnRH administration and have sought an explanati on. Some of the original evidence for the need for GnRH pulsatility was found in monkeys with hypothalamic lesions given exogenous GnRH (Belchetz et al., 1978). In that study, continuous administration of GnRH caused a sustained decline in p ituitary function, whereas intermittent administration restored normal pituitary func tion. Continuous GnRH infusion in ewes resulted in a peak LH response by 2 hours, which declined thereafter until reaching a
18 steady state at 20 hours (Nett et al., 1981). In the same st udy, FSH had a smaller peak by 2 hours and reached baseline by 5.5 hours. One possible mechanism for the declining LH response observed following continuous GnRH administration is the depletion of pituita ry gonadotropin stores. During the experiment by Nett et al. (1981), pituitary stores of LH /FSH were not affected and could not explain the declini ng release in response to c ontinuous GnRH infusion. In contrast to this finding, Clar ke et al. (1987) determined that LH pulse amplitude was correlated with pituitary stores, and that both amplitude and pituitary stores of LH were decreased when GnRH was given hourly comp ared to every 3 hours. In cattle, the continuous administration of GnRH cau sed a decrease in the amount of LH mRNA but did not affect FSH synthesis (Vizcarra et al., 1997). Furthermore, hourly administration of GnRH improved luteal activity and resulted in an increased number of large follicles by days 7, 9 and 11 when compared to either GnRH administered continuously or once every four hours (Vizcarra et al., 1997). Another possible mechanism for the dec lining pituitary response to continuous GnRH administration is the down-regulati on of GnRH-R caused by its own ligand, GnRH. In ewes continuously infused with GnRH there was an increase in GnRH-R numbers between 0 and 4 hours of continuous administration, but th ereafter receptors decreased until they were only half of thei r pre-infusion level by 24 hours (Nett et al., 1981). In vitro results, using cultured pituit ary cells treated with continuous or hourly GnRH pulses, there was a 12.8 fold increas e the GnRH-R mRNA in the hourly pulsed cells while there was no change in the conti nuously infused group (Kai ser et al., 1993). In cattle infused with continuous GnRH, there was a decrease in both the GnRH-R mRNA
19 and GnRH-R compared with cattle given pulsatile GnRH (Vizcarra et al., 1997). Frequency of pulsatile GnRH administration, given hourly or once every 3 hours, did not affect GnRH R numbers, even though LH pulse amplitude was decreased in the group receiving hourly GnRH (Clarke et al., 1987). Extrapituitary functions of GnRH GnRH receptors have been identified in various extrapituitary tissues including the ovary, placenta and prostate (Millar et al., 2004). Although many of the extrapituitary actions of GnRH are mediat ed indirectly by the gonadot ropins, LH and FSH, some studies have suggested a direct action of Gn RH. In females, long term administration of GnRH or a GnRH agonist has caused luteol ysis, decreased fetal survival, decreased serum estrogen and progesterone concentra tions, delayed puberty, delayed parturition, and inhibition of follicular maturation and ovul ation (Hsueh and Jones, 1981). In males, long term administration of GnRH has cause d decreased testicular androgen production, decreased LH and FSH receptor content, decreased testicular weight, inhibition of spermatogenesis, and decreased prostate and seminal vesicle growth (Huesh and Jones, 1981). Postulated mechanisms for the effect s of GnRH agonist administration listed above include; desensitizati on of gonadotropes to GnRH, de sensitization of gonadal cells to increased LH, or a direct extrapituitary action of GnRH (Hsueh and Jones, 1981). The latter possibility is supporte d by evidence indicating direct uptake of GnRH within the ovary (Hsueh and Jones, 1981). The luteolytic effect of GnRH agonist treatment seems to be species specific, such that in human and ra ts it is luteolytic (Huesh and Jones, 1981) and in ruminants it is luteot ropic (Davis et al., 2003). Th e positive effects of GnRH on
20 the ruminant CL are due to the luteotropic e ffects of LH and possibl y, the formation of an accessory CL. There is evidence that GnRH also ha s activity as a neurotransmitter at the hypothalamus and higher centers. GnRH neurons have synaptic contact with 5% of GnRH neurons projecting to the median eminen ce, indicating a role in regulating its own transmission (Leshin et al., 1988). Different st udies have also suggest ed that GnRH plays a role at higher brain center s in the expression of sexual be havior and estrus (Hsueh and Jones, 1981). Clinical Applications GnRH has been widely used in the cattle i ndustry because of its ability to induce an LH surge from the pituitary, its relatively small size, easy synthetic production, and lack of antigenicity compared to hCG. The GnRH induced LH surge causes ovulation of a dominant follicle (if present), CL formati on, and the subsequent emergence of a new follicular wave. Clinical applications of Gn RH include the treatment of cystic ovarian disease (Kittok et al., 1973), formation of an accessory CL as an aid for embryo survival (Schmitt et al., 1996), and synchronization of ovulation for timed insemination programs (Pursley et al., 1995; Momcil ovic et al., 1998). The underlying mechanism of cystic ova rian disease is a hypothalamic lesion involving the ER (Gumen and Wiltbank, 2002). This lesion can be corrected by sufficient exposure to progesterone which cau ses the up-regulation of the ER (Gumen and Wiltbank, 2002). An injection of GnRH in cystic cows causes ovulation and CL formation in 75-80 % of cases (Archbald a nd Thatcher, 1992; Farin and Estill, 1993). The progesterone exposure which occurs after CL formation corrects the hypothalamic lesion and leads to resumption of normal ovari an function (Gumen and Wiltbank, 2002).
21 In programs designed to synchronize ovulation, GnRH is used to ovulate a dominant follicle from two se parate waves. The first in jection causes ovulation of a dominant follicle, CL formation and the emergence of a new follicular wave. An injection of prostaglandin can then be us ed to lyse the CL on day 7, while a second injection of GnRH on day 9 causes ovulation of the dominant follicle which arose after the 1st injection of GnRH. Using this protocol the time of ovulation can be predicted and the cow can be inseminated without the de tection of estrus (P ursley et al., 1995; Momcilovic et al., 1998). Given that LH is the primary luteotropi n supporting CL function in the cow, it has been speculated that exogenous supplementa tion of GnRH could improve fertility in cows by increasing LH and subsequently, pr ogesterone production. Administration of GnRH every 2 hours for 72 hours caused an incr ease in luteal wei ght and progesterone content (Adams et al., 1975). A single inject ion of GnRH or agoni st can create an accessory CL which produces additional progest erone but does not necessarily improve pregnancy rates (Schmitt et al., 1996). It wa s then speculated that longer acting implants of GnRH agonists may improve luteal func tion by providing a greater duration of LH support. Longer acting GnRH agonist implan ts increased mean LH concentrations, increased progesterone concentration, and prol onged luteal lifespan, even in the absence of an accessory CL and regardless of whether the implant was given on day 3 or day 12 of the estrous cycle (Davis et al., 2003). The use of longer acting GnRH agonist within timed insemination programs has also been evaluated for its potential to increase CL function and inhibit follicular development. When a longer acting GnRH a gonist implant, Deslorelin, was used to
22 replace the second GnRH inject ion in a timed insemination protocol it caused minimal and sporadic increases in plasma progesterone, decreased follicular growth and decreased expression of estrus in respons e to a prostaglandin injection 16 days later (Bartolome et al., 2004). Another study used two different doses of Deslorelin implants (450ug and 740ug) in place of the second Gn RH of a timed insemination protocol, and compared pregnancy rates and progesterone concentr ation following treatment (Santos et al., 2004a). These results indicated that the pregna ncy rate in response to the lower dose was comparable to the standard protocol, while the higher dose had a decrease pregnancy rate. Folliculogenesis Folliculogenesis is the process by which primordial follicles, defined as an oocyte surrounded by a single layer of squamous preg ranulosa cells, are r ecruited and develop into antral follicles. They grow within a c ohort of developing follicles until one arises as the dominant follicle, which eventually either ovulates or becomes atretic (Richards, 1980). These follicular cohorts are referred to as waves and usually occur two or three times during the estrous cycle. In slaughter house ovaries from heifers on known days of the estrous cycle, follicles >5 mm showed tw o waves of growth, between the 3rd and 4th days and the 12th and 14th days. Each wave resulted in a preovulatory size follicle, the follicle from the first wave underwent atresi a while the follicle from the second wave ovulated (Rajakoski, 1960). Using ultrasonogr aphy, follicles can be mapped on the ovary and precisely followed throughout their grow th. When ten nulliparous heifers were followed by ultrasound, seven had three, and th ree had two follicular waves (Pierson and Ginther, 1988). In adult cows, there are generally two tran sient increases in the concentration of serum follicle-stimulating hormone (FSH): one coincides with the preovulatory LH surge
23 and one which occurs 12-24 hours after th e luteinizing hormone (LH) peak (Dobson, 1977). At the emergence of each follicular wave there is a 5075% increase in the concentration of serum FSH, and it has been suggested that this is the main initiating factor for a follicular wave (Adams et al., 1992). If the rise in FSH is delayed or inhibited, the follicular wave may be dela yed or inhibited (Fortune, 1994). Exogenous FSH can recruit greater than normal numbers of follicles and suppor t their growth beyond 5 mm. This leads to more follicles available for ovulation and is referred to as superovulation. These effects are dose-dep endent and small amounts of exogenous FSH can produce co-dominant follicles (Rivera and Fortune, 2001). Follicle-stimulating hormone may not be the only factor required for the initiation of a follicular wave since some hypophysectom ized sheep and other laboratory animals also exhibit recruitment of primordial follicles and limited growth of a reduced number of preantral follicles (Roy and Greenwald, 1989). Follicles less than 4 mm are not dependent on the acute support of gonadotropi ns (Gong et al., 1996). While increasing concentration of FSH is the in itiating cause of a follicular wa ve, it is the decline in FSH that signals the end of selection and development of a dominant follicle. All the growing follicles within a cohort equal to or greater than 5 mm contribute to the decline in concentrations of FSH (Gibbons et al., 1999) through the production of estradiol (Ginther et al., 2000a) and inhibin (Kaneko et al., 1995). The follicles in the developing cohort continue to require FSH for their growth and development even when the levels are declining. This was demonstrated by treati ng heifers with 6.0 mm growing follicles with estradiol (Ginther et al., 2000a). In this st udy, exogenous estradiol caused a decrease in FSH concentrations and delayed follicular development to the 10 mm stage by 48 hours
24 over controls. This study disproved the previ ous theory that the initial FSH surge was the major factor stimulating diameter growth in follicles from wave emergence to deviation. Evidence was also generated that the largest follicle during the growing phase was more sensitive to the decrease in FSH concentrati on. This was shown when the largest follicle lost its advantage and the second largest fo llicle overtook the largest to eventually become dominant in all treated animals compar ed to untreated contro ls (Ginther et al., 2000a). The point at which the future dominant follicle begins to surpass the others in growth and estradiol produc tion is known as deviation. Numerous studies are in agreement that morphologic deviation occurs when the largest foll icle is on average 8.5 mm in diameter (Beg at al ., 2002; Ginther et al., 1996; Ri vera and Fortune, 2003). The proposed close two-way functional coupling th at occurs between the dominant follicle and FSH concentrations states that when the follicular wave has reached the expected beginning of deviation, the low FSH concentrat ions are caused by the largest follicle, then are utilized and required by the largest fo llicle to establish domi nance (Ginther et al., 2000b). This was demonstrated by treating heif ers with estradiol to lower FSH levels when the largest follicle was 8.5 mm. The result was a transient decrease in FSH concentration and delayed growth of the follicle within 8 hours of the FSH nadir (Ginther et al., 2000b). The dominant follicle also has the greatest role in suppressing circulating FSH concentrations as shown in studies ab lating the largest follicle at 8.5 mm and demonstrating an increase in FSH compared to controls (Ginther et al., 2000a). The increase in FSH concentrations caused by ablati on occurs in less than 8 hours (Ginther et al., 1999).
25 After deviation, the low FSH concentrati ons are only required by the dominant follicle and are already too low to support de velopment of the smaller follicles, which cease growing and eventually regress (Ginther et al., 2000b). In an in-vitro study using cultured granulosa cells, the lowest of 3 doses of FSH stimulated the greatest production of estradiol, inhibin A, activin A and folls tatin, while the maximum progesterone output was achieved at the two higher doses of FSH (Glister et al ., 2001). These dose-dependent effects of FSH on granulosa cell products s eem to mimic in vivo actions. This occurs both at the time of deviation when FSH c oncentrations are low and the follicle is producing more estradiol, inhibin, activin and follistatin; and dur ing the preovulatory gonadotropin surge when FSH is higher, gr anulosa cell luteini zation occurs and progesterone output increases, while aromatase, inhibin/ activin a nd follstatin are very much decreased (Glis ter et al., 2001). The dominant follicle switches from FSH to LH dependency just after the point of deviation or >8.5 mm. Follicle-stimulating horm one is needed for follicular growth up to 9 mm, beyond this point LH is required fo r growth up to ovulatory size (Gong et al., 1996). A small elevation in LH lasting fo r about 48 hours occurs around the time of deviation (Ginther et al., 2001a). The signi ficance of the devia tion LH surge was examined by comparing the effects of progest erone treatment prior to and encompassing deviation, thereby inhibiting the normal incr ease in LH at this time (Ginther et al., 2001a). This resulted in no change in the ti ming of deviation, but caused a decrease in follicular fluid estradiol, oestrone, androstene dione, and free-IGF. However, there was an increase in follicular fluid IGFBP-2 with no change in follicular progesterone and immunoreactive inhibin. Changes in plasma were reflected as a decrease in estradiol
26 within 8 hours, an increase in FSH concentr ations, and the elimination of the typical decrease in FSH concentrations which occurs at deviation and accompanies the transient increase in LH. The increase in synthesis and se cretion of estradiol into circulation at the beginning of deviation is likely a function of the increase in LH. The reduced levels of oestrone and androste nedione in treated animals also indicated that LH stimulated increases in estradiol through the steriodoge nic pathway (Ginther et al., 2001a). The transient increase in LH around deviati on is important for the function of the follicle, but its absence does not prevent diameter growth and morphologic deviation. Luteinizing hormone is only critical for conti nued growth in the days following deviation (Ginther et al., 2001b). The importance of LH for growth was demonstrated when progesterone was given to lower LH concentr ations in preand post-deviation groups. There was no effect on the largest follicle in the pre-deviation group, while in the post deviation group the largest follic le was smaller than controls, had lower concentrations of estradiol, and had lower free IGF-1 concentr ation in follicular fluid (Ginther et al., 2001b). The theca interna cells of antral follic les typically express LH receptor (LHr) mRNA, and this expression increases with follic ular size (Xu et al., 1995). Despite this, there is no apparent relations hip between LH binding to the ca interna cells and follicle size. Expression of LHr mRNA is increased in the theca interna cells of dominant follicles (Xu et al., 1995). An important diffe rence between the largest follicle (F1) and the second largest follicle (F2) after deviation is th e expression of LHr mRNA in granulosa cells > 8mm (Xu et al., 1995). Gra nulosa cells had increased LHr activity on the day after deviation and an increase in the difference of LHr mRNA expression
27 between F1 and F2 8 hours prior to an increas e in diameter and estr adiol concentrations (Beg et al., 2002). Therefore, th e elevation in LH concentration and expression of LHr in granulosa cells are some of the first events leading to deviation. Follicle stimulating hormone (FSH) has been shown to induce th e expression of LHr mR NA in rat granulosa cells (Rani et al., 1981). Therefore, FSH responsiveness is crucial for the future responsiveness of the follicle to LH. The expression of LHr on granulosa cells may give the future dominant follicle an advantage over the others during the increase in LH accompanying deviation. The binding of LH to granulosa cells triggers the enzymatic processes involved in th e conversion of androgens into estradiol. Estrogen is a highly significant hormone during follicle growth and deviation. Some of its actions in the dominant follicle in clude; stimulating further follicular growth (Beg et al., 2003), stimulating proliferation of granulosa cells (Richards, 1980), a positive role in modulating the FSH dependent inducti on of LHr mRNA in granulosa cells (Rani et al., 1981), enhancing steriodogenesis, and increasing production of insulin-like growth factor 1 (IGF-I; Hsu and Hamm ond, 1987). At the point of deviat ion, there is a significant difference in follicular fluid estradiol concentrations, with the largest follicle having higher estradiol compared to the 2nd and 3r d largest follicles (Beg et al., 2002). The change in estradiol concentr ation and diameter between th e two largest follicles is generally the point that defines deviation. As early as Day 2 of a follicular wave, the future dominant follicle had hi gher concentrations of estrad iol in follicular fluid and granulosa cells secreted more es tradiol in culture than future subordinate follicles (Evans et al., 1997).
28 Another area under investigation for its ro le in the deviation phenomenon is the IGF system. This system includes insulin, insulin-like growth f actors I (IGF-I) and II (IGF-II), their binding proteins (IGFBP); specifically IGFBP-2, -3, -4 and -5, of which -2, -4 and -5 are most significant in the ovary; and the IGFBP proteas es. Insulin, IGF-I and IGF-II share 45% of the same amino acid se quences (Spicer and Echternkamp, 1995) and have similar actions at many of the same recep tors. IGF-I is produced in the liver and in the ovary by granulosa cells (S picer et al., 1993), but the majo rity present in the ovary is derived from peripheral circulation. IGF-II is primarily produced by theca cells and is less potent than IGF-I (Lucy, 2000). IGF-I is a pleotrophic growth factor that stimulates growth and development in a variety of ce ll types. In the ovary, IGF has a mitogenic effect stimulating an increase in granul osa cell (Spicer et al., 1993) and theca cell (Stewart et al., 1995) numbers. IGF has a significant role in steriodogenesi s, which is the basis for its contribution to deviation. There are three mechanisms wh ereby IGF increases steriodogenesis. Firstly, IGF directly activates ster iodogenic enzymes (Yang and Rajamahendran, 1998). Estradiol production increased when IGF was injected into follicular fluid (Ginther et al., 2004), or when pumped into the ovary (Spicer et al., 2000). Estradiol also increased when IGF was added to granulosa cell cultures (Spicer a nd Echternkamp, 1995). Se condly, IGF induces the expression of gonadotropin receptors. In cultured thecal cells, IGF increased the expression of LHr mRNA (Stewart et al., 1995 ). Thirdly, IGF has a synergistic effect with LH and FSH in the production of androge ns and estrogens. In the presence of LH, IGF-I increased androstenedione and progesterone production in cultured thecal cells, whereas LH or IGF alone had li ttle or no effect (Stewart et al., 1995). In granulosa cells,
29 the addition of IGF-I enhanced the effects of low doses of FSH on secretion of estradiol (Glister et al., 2001), and this effect was not dose dependent, where a lower dose of IGF (100ng/ml) increased estradiol more than a hi gher dose of IGF (200ng/ml; Spicer et al., 1993). In the same study, progesterone producti on was also greater in granulosa cells treated with IGF and FSH than with IGF alone. Total IGF levels are often re ported to be the same in the dominant and subordinate follicle. The primary difference is in the leve l of free vs. bound IGF. Dominant and future dominant follicles consistently have higher levels of free IG F and lower levels of the low molecular weight IGF binding proteins than the subordinate follicl e (Lucy, 2000; Beg et al., 2002). After ablation of the dominant follicl e, two of the earlies t changes during the assumption of dominance for F2 were an increase in free IGF-I and a decrease in IGFBP2 (Beg et al., 2002). This occurred when the follicle was in the diameter range of 8.4 8.7 mm. Another study found that the most consistent predictors of future dominance in 5 8 mm follicles were high follicular fluid concentrat ions of estradiol and low concentrations of IGFBP-4 (Mihm et al., 2000). In contrast to the previously menti oned studies, Mihm et al. (2000) found that IGFBP-2 was not di fferent between future dominant and subordinate follicles. This could be due to the earlier developmental stage of the follicles in this study (5 8 mm vs. > 8 mm). Anot her possible explanation for the contrast between these studies could be accuracy of the tests used to measure and distinguish between the binding proteins. However, the impo rtant point is that the follicle destined for dominance has fewer low molecular weight IGF binding proteins; primarily IGFBP-4. The mechanism through which the follicle develops a decreased number of low molecular weight IGFBP and increased concen trations of free IGF may be the critical
30 aspect to the establishment of dominance. There are three primary possibilities through which a decreased amount of IGFBP-4 may occur. These possibilities include 1) decreased genetic expression for IGFBP-4, 2) alterations in proteo lytic degradation, and 3) decreased uptake from peri pheral circulation. The majority of research has focused on the first two explanations, and the answer may be a combinati on of all three. In healthy follicles, theca cells continually expre ss mRNA coding for IGFBP-4 while granulosa cells express mRNA for IGFB P-2, but only up to a diameter of 8 mm (Armstrong et al., 1998). Since expression of mRNA for IGFBP-4 did not change while its concentrations decreased, the authors postulated that its c ontrol is not through genetic expression but possibly by a protease. The change in expression for IGFBP-2 suggests that its control is genetic and was inhibited by increasing FSH (Armstrong et al., 1998). Interestingly, IGFBP-4 was found in granulosa cells after dete rmining that these cells do not express its mRNA, and this lead to the conclusion that IGFBP-4 was brought into the cells from peripheral circulation (Armst rong et al., 1998). It is likel y that IGFBP-4 acts as a transport and storage unit of available IGF for granulosa cells. Free IGF can then be accessed by activation of the specific prot ease. Another study found that IGFBP-4 is degraded by a metallodependent protease (Mazerbourg et al., 2000) which was later identified as pregnancy associated plasma protein A (PAPP-A; Rivera and Fortune, 2003). Furthermore, IGFBP-2 is detrimental to development through sequestration of IGF which is needed for growth and de velopment (Armstrong et al., 1998). The proteolytic activity involved in the degradation of IGFBP-4 was enhanced when IGF was added to follicular fluid, and decreased when IGFBP-2 and -5 were added (Mazerbourg et al., 2000). This agreed with Rivera and Fortune (2003) who demonstrated
31 a greater affinity of the proteolytic enzy me for IGFBP-4 when bound to IGF. Follicle stimulating hormone induced co-dominant fo llicles which showed similar proteolytic activity against IGFBP-4 as normal dominant follicles and greater than that of subordinate follicles (Rivera and Fortune, 2001) This suggested a potential role for FSH in inducing the protease, but di d not rule out the simultaneous increase in estradiol as a factor. This was later addre ssed by the same authors who found that the increase in IGFBP-4 proteolytic activity and a decrease in IGFBP-4 we re the earliest changes in normal growing follicles and FSH induced co -dominant follicles, occurring prior to significant changes in estradio l concentration or diameter (Rivera and Fortune, 2003). Cystic Ovarian Disease in the Dairy Cow Introduction Cystic ovarian disease is a well-recognized condition in dairy and beef cattle. It occurs more commonly in cows than in heif ers and is a signifi cant source of economic loss for dairy farmers. It is a condition in which cows are anovular for an extended period of time, usually in the first 60 days po stpartum period, even though a preovulatory follicle exists on the ovary. The exact underlying endocrinology is not completely understood, but it appears to be related to an unresponsiveness of the ER in the hypothalamus to follicular estrogen at estrus (Gumen and Wiltbank, 2002). Ovarian cysts can be classified in three ways : follicular cyst, luteal cyst or a cystic corpus luteum (Roberts, 1971). A follicular cyst is an anovulatory follicle, and the pathology and treatment will be th e focus of this section. A lu teal cyst is an anovulatory follicle that became partially luteinized. A cy stic corpus luteum is a functional corpus luteum that developed a cavity, without disruption of the estrus cycle. The first two types
32 arise from the same pathology and disrupt nor mal cyclicity, while the third type is not considered abnormal (Roberts, 1971). Definition and Diagnosis The definition and diagnostic criteria for ovarian cysts are con tinually evolving as is evident by the inconsistency and change s over time in the literature. The classic definition for ovarian cysts are multiple or single follicle like structures, greater than 25 mm and persisting for 10 days or longer in the absence of a CL (Bie rschwal et al., 1975; Kesler and Gaverick, 1982; Fari n and Estill, 1993). In the more recent literature, they are defined as a follicle that persists for more th an 6 days (Lopez-Gatius et al., 2002; Silvia et al., 2002) and has a diameter e qual to or greater than ovulat ory size (> 17 mm; Silvia et al., 2002; Halter et al., 2003). The other importa nt criterion for a folli cular cyst is the absence of luteal tissue (Silvia et al., 2002) or a corpus luteum (Lopez-Gatius et al., 2002). Diagnosis of an ovarian cyst can be achie ved by palpation per rectum, a history of anestrus or nymphomania, ul trasonographic examination a nd serum progesterone levels. Findings per rectum will include a flaccid uter us similar to diestrus, mucometra if the condition is chronic (Youngquist, 1986), single or multiple smooth, fluid-filled structures greater than 17 mm, and the ab sence of a CL. Differential diagnoses for ovarian cysts diagnosed by palpation per rectum includes a normal preovulatory follicle, a CL, a corpus hemorhagicum, adhesions, salpingitis, hydrosalpinx, oophoritis, ovarian abscess, neoplasia, and cysts of the fimbria (Youngquist, 1986). Palpation of the uterus per rectum and ultrasonographic examination of the ovari es and uterus will aid in differentiating these structures. Ultrasonographic examination of the ovaries will reveal that the cyst is thin-walled and contains an anechoic antrum measuring greater than 17 mm. A luteinized
33 cyst will have increased wall thickness and feel firm on palpation per rectum, and on ultrasonographic examination will often have gr ay echogenic patches along the inner cyst wall or within the antrum (Farin et al., 1990) and a wall greater than 3 mm thick (Ribadu et al., 1994). A normal corpus luteum has an echogenicity different than the surrounding tissues and a well defined border (Pierson and Ginther, 1984). Many normal corpora lutea contain an anechoic central cavity (cystic CL). These have been observed temporarily from day 5-7 of the estrous cycle in a proportion of heifers examined daily (Pierson and Ginthe r, 1984) and in approximately 30% of slaughterhouse ovaries examined (Peter, 1997). Ultrasonographic examination of the uterus would reveal signs typical of diestrus (Pierson and Ginther, 1988). If the uterus has tonicity, responds to palpation, and ultrasonogr aphy reveals a heterogenous texture, these are sign of impending estrus (Pierson and Gi nther, 1988), and help differentiate between a follicular cyst and preovulatory follicle. The accuracy of palpation per rectum, u ltrasonography and peripheral progesterone levels to diagnose ovarian cysts has been deba ted in the literature. Palpation per rectum alone was reasonably accurate, but varies be tween studies and cyst type. One study found that when palpation per rectum was used to di fferentiate between cysts, whether follicular or luteal, and other ovarian structures the accuracy was 52% (Ribadu et al., 1994). Most luteal cysts can be successfully treated with prostaglandin. Therefore, it is generally beneficial to differentiate between follicular and luteal cysts, alt hough this is where most of the inaccuracy occurs. Spr echer et al., (1988) found that palpation per rectum had a 75% positive predictive value (PPV) for follicular cysts, with a sensitivity and specificity of 61.9% and 50%; while for luteal cysts it ha d a PPV, sensitivity a nd specificity of 31%,
34 50.0% and 61.9%, respectively. In a different st udy, diagnosis of luteal cysts by palpation per rectum had a sensitivity, specif icity and PPV of 43.4%, 64.7% and 68.4%, respectively; while ultrasonography had a se nsitivity, specificity and PPV of 86.7%, 82.5% and 89.7%, respectively (Farin et al., 199 2). For follicular and luteal cysts, one study found that the milk progesterone en zyme immunoassay had a PPV of 91.8% and 83.3% respectively (Sprecher et al., 1988). On average, the PPV for follicular cysts diagnosed by rectal palpation or ultrasonogr aphy are 66% and 74%, respectively; and the PPV for luteal cysts diagnosed by palpation per rectum or ultras onography are 66% and 85%, respectively (Hanzen et al., 2000). Overall, these results indicate that 1) pa lpation per rectum is more accurate for diagnosing follicular cysts over luteal cysts (Sprecher et al., 1988), 2) ultrasonography was better than palpation per rectum, especially for detecting luteal cysts (Farin et al., 1992) and 3) palpation per rectum combined with progesterone leve ls or ultrasound is generally more accurate than palpation per re ctum alone (Sprecher et al., 1988, Farin et al., 1990). The wide range of sensitivity, specificity and positive predictive values in the literature is partially due to the range in criteri a used to define follicular and luteal cysts. After a cyst is diagnosed by palpation per rectum or ultr asonography, most studies use progesterone levels as the gold standard to differentiate between follicular and luteal cysts. Progesterone values used to determin e a luteal cyst range from greater than 0.5ng/ml (Farin et al., 1990) to greater th an 5.0ng/ml (Sprecher et al., 1988). Some studies have found intermediate to high proge sterone levels when follicular cysts were diagnosed by ultrasound as having no luteal tissue (Jeffcoate a nd Ayliffe, 1995) or low
35 progesterone when luteal cysts are diagnosed as having wall thickness greater than 3mm (Douthwaite and Dobson, 2000). A possible expl anation for the luteinized cyst with progesterone levels <0.9ng/ml is that it has a function similar to a late CL, around day 19, when it can be observed but secretes very l ittle progesterone and is unresponsive to PGF (Douthwaite and Dobson, 2000). Ther e can be considerable varia tion in the density of the luteinization as measured by ultrasound (Farin et al., 1992) and occasionally histology is the only method that can accurately determine th e presence of luteal tissue (Cook et al., 1991). A low level of peripheral progester one does not distinguish between a preovulatory follicle and follicular cyst (Douthwaite and Dobson, 2000). There are numerous ways to accurately diagnose a folli cular cyst. Probably the preferred approach is a combination of progesterone, ultrasonograp hy and histology. It appears that the most practical method for field diagnosis relies on car eful palpation per rectum of the ovaries and uterus, complimented by ultrasonogra phy whenever possible or a follow-up examination in one to two weeks. Incidence Rate and Risk Factors The incidence of cystic ovarian disease is reported in the literature to be between 7.7 and 17% for dairy cows (Table 1) with an overall mean of 11.9%. The reported incidence on a given farm may depend on th e frequency of examinations, where more frequent visits and a shorter in terval (Erb and White, 1981) or a very long interval (>30d) where cows have time to spontaneously cure before being detected could both result in a lower incidence rate. Most i nvestigations report a peak incidence rate during early lactation (Table 2) while a few have found an additional peak occurring between 150-220 days, giving a bimodal distri bution. A possible explanation fo r the second peak is an
36 increased intensity in examinations of non-pr egnant cows at this time (Bartlett et al., 1986). Table 1. Reported incidence rates for cyst ic ovarian disease in the literature. Literature cited # Lactations % Lactations Morrow et al., 1966 357 12.3 Bierschwal, 1966 1 436 13.0 (11.1Holstein, 17.0Guernsey) Whitmore et al., 1974 375 11.2 Erb and White, 1981 1 599 12.4 Bartlett et al., 1986 2 847 12.8 Grhn et al., 1998 7 523 10.6 Fleischer et al., 2001 2 197 11.7 Hooijer et al., 2001 15 562 7.7 LopezGatius et al., 2002 873 13.1 (43-49d) 11.2 (57-63d) Table 2. Reported stages of lactation with an increased inciden ce of cystic ovarian disease. Many investigations have found varying trends in incidence, given certain characteristics or risk factors. These risk factors may provide useful insight into the underlying etiology of cystic ov arian disease. Two main risk factors that have been investigated are the effects of milk production and lacta tion number or parity. The incidence of cystic follicles was less during th e first lactation compared with the second and third lactations (Whitmore et al., 1974; H ooijer et al., 2001). Ther e is also evidence that there is an increasing inci dence of cystic follicles with increasing age (Bartlett et al., 1986). Milk production and lacta tion number have been ne gatively correlated with spontaneous early cyst recovery (Lopez-Gati us et al., 2002). Lact ations with cystic follicles were associated with 422 kg more 305 ME-D milk production than lactations Literature cited # Lactations First Peak Second Peak Morrow et al., 1966 357 1st post-partum ovulation Not observed Bierschwal, 1966 1 436 Days 2160 Not observed Whitmore et al., 1974 375 Days 16 to 45 Not observed Erb and White, 1981 1 599 Days 31-60 Days 151180 Bartlett et al., 1986 2 847 Days 31-40 Days 190-220
37 without cystic follicles (Bartle tt et al., 1986), and a 1 kg increase in milk yield resulted in a 1.05 increased risk for cysts (Lopez-Gatius et al., 2002). Another study found that cows with follicular cysts produced an average of 379 kg more milk in the first 90 days post partum, and 438 kg more by 305 days than thei r non-cystic herd-mates (Johnson et al., 1966). Cows experiencing abortion, dystocia, twins, retained fetal membranes, metritis, ketosis or other debilitating disease had a sign ificantly higher incidence of cystic follicles compared to cows not experiencing these conditions (Morrow et al., 1966). Another study found that cows with an abnormal puerperium had a 1.9 times higher risk for developing cystic follicles than normal cows (Lopez-Gatius et al., 2002). Cows which became lame during the first 30 days postpartu m had a significantly higher incidence of ovarian cysts compared to non-lame contro ls (25.0% vs.11.1%, respectively; Melendez et. al., 2003). There is evidence for a seasonal increase in the incidence of cysts in some areas. In one study, cows calving in the summ er months were 2.6 times more likely to develop cysts than those calvi ng in the winter (Lopez-Gat ius et al., 2002). However, other authors found no significant effect of season (Melendez et al., 2003; Bartlett et al., 1986). Etiology A genetic susceptibility to cystic ovarian disease has been suggested and researched by a number of authors. In one case study, a closed 300 cow dairy farm using natural service found that two of thei r eleven bulls produced a sign ificantly higher proportion of daughters with cystic ovaries than expected, suggesting a he ritable basis (Kirk et al., 1982). In recent studies using computerized reco rds, the estimated heritability of cystic
38 ovarian disease has been determined to be 0.0870.102 (Hooijer et al., 2001) and 0.050.08 (Zwald et al., 2004a). In a study using producer recorded data, th e estimated genetic correlations between cysts and ketosis was 0.42, cysts and lameness was 0.16 and between cysts and displaced abomasum was 0.17 (Zwald et al., 2004b). The in cidence of cystic ovarian disease also has a positive correlation with milk production tr aits, especially to protein levels (Hooijer et al., 2001). Therefore, selec tion pressure for milk producti on traits may inadvertently increase the incidence of COD over time. In one example, it was estimated that a 500 kg increase in 305-day milk yield will increas e the incidence of C OD by 1.5% (Hooijer et al., 2001). The possibility of increased suscep tibility due to breed was suggested in a study which found an increased incidence in Guernseys, 18% vs 11% in Holstein Friesians (Bierschwal, 1966). Peripartuient disease has been linked to ovarian cysts, and several studies have found a positive relationship between the inci dence of metritis and ovarian cysts (Fleischer et al., 2001; Morrow et al., 1966; Lopez-Gatius et al., 2002). A possible explanation for the basis of metritis leading to ovarian cy sts is through the release of endotoxin by pathogenic gram-negative bacteria present in the uterus. Evidence for the role of bacteria and endotoxin was revealed during a comparison of postpartum cattle in which those who later developed cysts ha d increased cortisol and PGF metabolites (PGFM) concentrations and higher intrauterine culture scores prior to detection of the cysts (Bosu and Peter, 1987). It has been shown that intraute rine infusion of endotoxin in heifers caused suppression of the LH surge, an d resulted in the formation of follicular cysts which persisted for 7 to 21 days (Peter et al., 1989).
39 The two primary sites where endotoxin is thought to exert its effects are the hypothalamus and the ovary. In the ewe, in travenous endotoxin infusion interrupted rising estradiol levels and delayed or inhibited the pr eovulatory LH surge. The interruption of the rising estradiol levels wa s due to an effect on follicular development (Battaglia et al., 2000). Furthermore, the eff ects of endotoxin on the estradiol-induced LH surge only occurred if the animal was expos ed to endotoxin during the first 14 hours of the estradiol signal, or during the estradio l reading phase (Battaglia et al., 1999). Endotoxin also acts at the hypothalamus to stimulate ACTH release which resulted in increased adrenal cortisol production (M oberg, 1971). The intrauterine infusion of endotoxin in the experiment by Pe ter et al. (1989) caused a sign ificant increase in cortisol levels. When ewes were infused with endotoxi n, those responding with a greater increase in coritsol and progesterone levels were also more likely to have an inhibited or delayed LH surge (Battaglia et al., 2000). Cortisol is a well-known inhibitor of the Gn RH induced LH releas e at the level of the pituitary. Dairy heifers tr eated with ACTH had a significantly depressed LH response when injected with LHRH (Matteri and M oberg, 1982). Heifers continually infused with ACTH had significantly decreased LH con centrations which remained decreased throughout the cycle, and cons equently, did not have an LH surge (Li and Wagner, 1983a). When repeated acute stress was app lied to cattle, a proportion of the stressed animals did not have an LH surge while all of the control animals had an LH surge. These stressed animals also experienced increased cortisol which decreased with subsequent stress periods (Stoebel and Moberg, 1982). In vitro studies with bovine pituitary cells showed that cortisol pretreatment drepresse d the LH response to GnRH (Li and Wagner,
40 1983b) and the cortisol concentrations typical of postpartum suckled cattle are sufficient to inhibit the GnRH induced LH surge (Padmanabhan et al., 1983). Given this information, the relationship be tween metritis and cysts is most likely due to direct actions of endotoxin on the ova ry and, indirectly, th rough the actions of cortisol induced by endotoxin. It is also through cortisol th at cattle under stress may be more likely to develop ovarian cysts, ma king stress another possible etiology for the disease. Fate and Endocrinology Bovine ovarian cysts are not static struct ures since they can persist for variable periods of time, or regress and be replaced by another cyst or an ovulatory follicle (Cook et al., 1990; Hamilton et al., 1995; Wiltbank et al., 2002). These three responses can be simplified into persistence, turnover or spontaneous recovery. A study which marked cysts with charcoal to follow their devel opment found that in no case the cysts ovulated (Cook et al., 1990). In this study, cysts either persisted for the duration of the study period (40 days) or regressed and were re placed by an upcoming follicle. The interval between the detection of a cyst and the next follicular wave was longer by a difference of 4-5 days, and more variable, in cystic cows than in normal cows (Hamilton et al., 1995). The endocrinology of cystic cows has been compared to normal cyclic cows by many authors (Zaied et al., 1981; Hamilton et al., 1995; Cook et al., 1991; Silvia et al., 2002; Wiltbank et al., 2002). There is no diffe rence in FSH concentrations or profile between cystic and normal cows (Hamilton et al., 1995; Cook et al., 1991), and an increase in FSH precedes each new follicular cyst wave (Hamilton et al., 1995). Cows with cystic follicles have higher basal LH c oncentrations during the follicular phase than normal cows, primarily due to an increased pulse frequency and amplitude (Hamilton et
41 al., 1995; Cook et al., 1991). These cows fail to have an LH surge at the expected time of ovulation which results in a continuously growing or persistent, anovulatory follicle (Hamilton et al., 1995). When GnRH is exogenously administered to cystic cows, the p ituitary consistently responds with an LH surge, resulting in lutein ization of the cyst and/or any other follicle present on the ovary with the ability to re spond to an LH surge (Kittok et al., 1973; Cantley et al., 1975). In most cases, after lu teinization occurs and progesterone levels rise, LH profiles return to normal and cyclicity resumes. Many cystic cows lack a GnRH/ LH surge in response to elevated endoge nous estradiol (Hamilton et al., 1995), or exogenously administered estradiol (Dobson a nd Alam, 1987; Refsal et al., 1988; Gumen et al., 2002; Gumen and Wiltbank, 2002). A comparison of the pituitary and hypothalamic concentrations of GnRH found that GnRH was lower in the combined preoptic area and hypothalamus proper in cyst ic cows (Cook et al., 1991). In the same study, anterior pituitary concentrations of LH, FSH, and receptors for GnRH did not differ between normal and cystic cows sugge sting that the functi on of the hypothalamus may be altered in cystic cows. Cows with cysts have variable serum and follicular fluid levels of progesterone and estradiol (Cook et al., 1990; Hamilton et al., 1995). These levels depend on the steriodogenic capacity a nd cell types present as well as th e proportion of luteinization in the cyst (Cook et al., 1990). In general, es tradiol concentrations were not different between cows with and without cyst s (Hamilton et al., 1995; Cook, 1991). When estradiol levels were evaluated based on follicul ar stage, at the point where the follicle/ cyst had reached ovulatory size, those which developed into follicular cysts had greater
42 estradiol concentrations than follicles whic h ovulated (Yoshioka et al., 1996; Hamilton et al., 1995). In ewes with induced persistent follicles, gonadotropin support in the form of LH pulses was required for follicular estrogen production for 10 days, at which point the follicle regressed and the ability to produ ce androstenedione and estradiol decreased despite continued LH pulses (Dobson et al ., 1997). This may explain why cows with follicular cysts may initially have elevated estrogen levels, which then decreases despite high mean LH concentration. Because they have no corpus luteum, cystic cows have lower progesterone concentrations than normal cyclic cows. Cys tic cows frequently have an intermediate level of progesterone, defined as 0.1-1.0 ng/ ml (Hamilton et al., 1995; Yoshioka et al., 1996; Cook et al., 1991). In one study, 66% of cows diagnosed with follicular cysts had intermediate levels of progesterone (Silvia et al., 2002). When monitoring cows with intermediate levels of progesterone, only 10% of developing follicles ovulated, and 66% resulted in cyst formation (Halter et al., 2003; Silvia et al., 2002). Meanwhile, cystic cows with low or high progesterone were more likely to ovulate or undergo follicular atresia than cows with intermediate progesterone (Ha lter et al., 2003). Pathogenesis The underlying pathogenesis involved in form ation of follicular cysts has not been definitively ascertained. The prevailing hypothe sis involves a lesion at the hypothalamus, in the area responsible for converting the pos itive feedback of es trogen into a GnRH surge and subsequent LH surge from the anterior pituitary. In the normal sequence of events, rising le vels of estradiol du ring the follicular phase, aided by declining levels of progester one, result in the hypothalamic GnRH surge. Administration of exogenous estradiol to cows with naturally occurr ing cysts has resulted
43 in a variety of responses. In general, th e normal and expected response of a timely estradiol induced GnRH and LH surge has b een disrupted in cows with cysts with evidence for it being delayed or absent. One study recorded that cows with follicular cysts had a delayed GnRH/ LH response to ex ogenously administered estradiol benzoate by approximately 8 hours (Zaied et al., 1981). In contrast to the majority of studies, chronically cystic cows which were later ovariectomized responded to estrogen challenge 6 weeks after ovariectomy with higher peak LH levels than normal cows (De Silva and Reeves, 1988). The robust LH response in th ese cows may have been due to the time elapsed and/or the ovariectomy which ma y have allowed time for recovery of hypothalamic function. In another study of natura lly occurring cysts, 47% of cows with luteal cysts (milk P4 >10 ng/ml) and 48% of cows with follicular cysts had a normal response to estradiol benzoate after treatment with cloprostenol (Nanda et al., 1991). The interesting aspect of this study was the si milar response in cows diagnosed with both types of cysts, lending evidence to the hypothe sis that both luteal and follicular cysts share the same underlying pathophysiology. Anot her study using 6 cows with low steroid concentrations and 6 under estrogenic influe nce found that administration of exogenous estradiol resulted in no LH response in 11/12 cows (Refsal et al., 1988). Cystic cows typically have increased basal serum LH c oncentrations (Cook et al., 1991; Hamilton et al., 1995). Elevated LH levels are important for the maintenance of the cysts, but may or may not be involved in th eir development. In order to determine if elevated LH concentration can cause cysts, Hampton et al. (2003) administered high, frequent doses of LH to cows on Day 1 after emergence of a follicular wave and measured follicular response. Cows receivi ng exogenous LH were not at an increased
44 risk of developing cysts compared to saline infused cows. These cows still had an LH surge and ovulated, despite low progester one and elevated LH levels. This study confirmed that although elevated basal LH le vels may be involved in the maintenance of cysts, it is not likely a fa ctor in the pathogenesis. The potential role of steroidogenic enzy mes, FSH and LH receptors has been examined for their role in the pathogenesis of cysts. The granulosa cells of chronic cysts have increased mRNA expression of LH receptors and 3 -hydroxysteroid dehydrogenase (3 -HSD) compared to normal dominant foll icles (Calder et al., 2001). These finding were speculated to be secondary to the hypothalamic-pituitary dysfunction causing low progesterone and increased LH c oncentrations typically of cyst s and not a direct cause of cyst development. A similar study comparing 3 HSD expression in granulosa and theca cells of normal, atretic and cystic follicles found that cysts had hi gher enzymes levels in granulosa cells and lower levels in theca cells compared to the normal follicles (Isobe et al., 2003). Compared to atretic fol licles, cysts had much lower 3 HSD expression in theca cells which was speculated to be a critical enzymatic ch ange necessary for atresia. This difference may be the cause of cyst pe rsistence and altered steroid production (Isobe et al., 2003). The lack of the positive feedback of estr adiol on LH release can lead to ovarian cysts. This was shown in a study immunizing cattle against estradio l then monitoring LH levels and follicular development (Kane ko et al., 2002). All of the cattle immunized against estradiol developed follicular cysts. They had a mean LH concentration higher than non-immunized cows, and did not have a preovulatory LH surge.
45 Similar endocrinological states and ovarian pathologies to cows with cystic ovarin disease have been induced in genetically modified mice with specific estrogen and progesterone receptor genes knocked out. Estrogen receptor alpha knock out ( ERKO) mice typically have hemorrhagic polycystic ova ries, elevated LH concentrations due to the lack of ER/ estrogen mediated ne gative feedback on the hypothalamic-pituitary axis, and an amplified steroidogenic pathway sim ilar to the follicular stage (Couse et al., 2004). The ER is the primary estrogen receptor present in the hypothalamus while the ER is the primary estrogen receptor present in the ovary and uterus (Couse and Korach, 1999). Estrogen receptor beta knock out ( ERKO) mice do not have hemorrhagic cystic ovaries, even in the presence of abnormally high pulsatile LH. During periods of high pulsitle LH, ERKO mice have a steroidogenic profile similar to the luteal phase (Couse et al., 2004). These re sults indicate that ER is needed for the negative feedback of estradiol on the hypothalamus and that the ER is needed for the phenotypical manifestation of cystic ovaries. Progesterone receptor knock out (PRKO) mi ce have a similar endocrinological and physical profile to the ERKO mice, with high basal LH levels and the lack of a preovulatory LH surge (Chappe ll et al., 1997). Although, one diffe rence is that follicles of the PRKO mice grow to ovulator y size, but not beyond as in the ERKO mice. PRKO mice lack the ability to produce an estrogen induced LH surge (Chappell et al., 1999). This indicates that the progesterone recepto r (PR) is required for transmission of the estrogen-induced signals leading to the LH su rge. In rats, a PR antagonist effectively blocked the estrogen-induced GnRH surge (Chappell and Levine, 2000). In the ewe,
46 progesterone priming increased the amplitude of the GnRH surge in response to estradiol (Caraty and Skinner, 1999) Localization of the specific area involved in transmitting the estradiol signal determined that the PRs located in the anterove ntral periventricular nu cleus are critical to transmission of the signal in the rat (Cha ppell and Levine, 2000). Progesterone receptors are up-regulated by estrogen (Chabert-Buffet et al, 2000). This was shown in the rat when estrogen induced PR expression within 2 hour s in a dose-dependent manner (Shughrue et al., 1997). The progesterone receptors induced by estrogen may even be transactivated in the absence of ligand (Levine, 1997). In this way, PR expression is induced by estrogen and then used as a transcript ional regulator in transmission of the estrogen signal to the GnRH releasing cells in the hypothalamus. Th e ligand independent activation of the PR likely mediates the neurosecretory estrogen feed back signals that result in an increased pituitary response to GnRH (Levine, 1997). Other aspects of uncovering the details of the estradiol induced LH surge involved determining the specific location of these receptors in the hypothalamus in various species and the factors controlling their expre ssion. Early studies in the rat localized ER mRNA expression in the arcuate nucleus and the ventrolateral portion of the ventromedial hypothalamus as areas under hormonal regulation (Simerly and Young, 1991). In this study, administration of estradio l to female rats for 24 hours caused a 40% down-regulation of ER mRNA with slightly more of a decrease occurring in the arcuate nucleus than in the ventromedial hypothala mus. In the ewe, ER staining localized expression to the preoptico-hypothalamic continuu m with areas of greatest density at the arcuate nucleus and the ventromedial hypothalamus (Blache et al., 1994).
47 Hormonal regulation of ER expression wa s examined through artificially-induced stages of the estrous cycle which demonstrat ed that, similar to the rat, estrogen caused down-regulation of ER expression, but also that mid-luteal phase or elevated progesterone concentrations caused an increas e in ER expression in the ventromedial hypothalamus (Blache et al., 1994). In this study, ER levels did not change significantly in the arcuate nucleus with various hormonal treatments. Since it is known that GnRH neurons contain few or no estrogen receptors, estrogen must activate ER-containing interneurons to transmit the signal to GnRH containing neurons (Caraty et al., 1998; Peterson et al ., 2003). In the ewe, the hypothalamic areas involved in estrogen f eedback regulation on the LH pulse were determined in a study using estradiol microi mplants (Caraty et al., 1998). These authors found that estrogen exerted str ong negative feedback effect on the medial preoptic area and caudal medial basal hypothalamus while it exerted a strong positive effect in the ventromedial nucleus. Collectively, these resu lts lend evidence to the hypothesis that estrogen transmits its signal to the GnRH containing neurons through ER-containing interneurons located in the ventromedial nucleus and possibly through progesterone receptors expressed in the medial preoptic area and anterior ventral periventricular nucleus. Estrogen is responsible for the down-regulation of its own receptor while progesterone is involved in up-regulating ER expression. It is postulated that dairy cattle with ovarian cysts have a compromised expression of ER in the hypothalamus and are unable to produce an LH surge in response to estrogen. Although ER levels in the hypothalamu s of cows with cysts have not been directly measured, there has been research on receptor expression in the anterior pituitary
48 and ovary. The abnormality identified between the ER and PR in the ovary and the ER and PR in the pituitary of cystic cows is th at they are not positively correlated, as would be expected in normal cows (Odore et al., 1999 ). Cows with cysts had similar expression of ER in the pituitary when compared to cows with normal dominant follicles, but the ovarian ER levels were much lower than in cows with normal dominant follicle and similar to levels expected during the luteal phase. In general, pituitary and ovarian receptor concentrations are correlated with each other and match with estrogen concentrations. Therefore, when estrogen leve ls are high, ovarian a nd pituitary receptor levels will also be elevated. Although cows with cysts had higher estrogen levels than normal cows, they had decreased expression of ER in the ovary (Odore et al., 1999). In a study examining the ER and PR expression in anestrous ewes, GnRH administration with and without progesterone caused an increase in pituitary ER and PR levels (Tasende et al., 2002). They also found that progesterone tr eatment decreased ER and PR levels in the uterus and that, in general, the receptor expression at the pituitary and uterus are correlated. Treatment with progesterone has been shown to correct the hypothalamic lesion that causes cows and ewes with cysts to be unable to produce an LH surge in response to estrogen. This likely occurs through the pr ogesterone mediated up-regulation of ER expression in the ventromedial hypothalamu s, and possibly thr ough activation of PR containing interneurons invol ved in the transmission of the estrogen signal to GnRH neurons. Support for this hypothesis is evident in studies where cows with cysts were unresponsive to estrogen treatment until exposur e to progesterone occurred for a period of 7 days (Gumen and Wiltbank, 2002; Nanda et al., 1991). The same response was seen
49 in ewes in which cysts were induced using constant estrogen admi nistration (Ozturk et al., 1998). Following this period, progesterone administration resulted in a normal response to subsequent estradiol challenge. According to the literature on mice, rats and ewes, abnormally high or abnormally low levels of estrogen preceding the expected time of ovulation could lead to the downregulation of its own receptor, as well as possibly the progesterone receptor, which are both involved in the transmission of the estr ogen signal. A GnRH/ LH surge in the absence of an ovulatory follicle and without s ubsequent progesterone exposure will result in the development of a follicular cyst (Gumen et al., 2002). This is likely because the hypothalamus requires progesterone exposure to up-regulate the ER and reinitiate the next signal. Any condition which may di srupt the timely GnRH/ LH surge could potentially lead to the deve lopment of ovarian cysts. Research examining the endocrinology of subfertile cattle has lead to the hypothesis that these cattle, wh ich are subfertile for variou s reasons, may share a similar lesion to cows with spontaneous ly occurring cysts. Early in vestigations comparing cows with cysts, lame cows, and thin, non-cyclic cows revealed that all three groups had the same lack of an LH response to exogenous ly administered estr adiol (Dobson and Alam, 1987). As an explanation for the thin, non-cyc lic cows, estradiol may have an especially strong negative feedback effect on the hypothala mus. It is postulated that in thin, postpartum cows experiencing negative ener gy balance, the hypothalamus is extremely sensitive to the inhibitory effects of estr adiol causing decreased LH secretion (Wiltbank et al., 2002). Cattle in a negative energy balance will, therefore, have follicular growth to the point of deviation, but as the follicle produces more estradiol, the hypothalamus
50 responds by decreasing GnRH and LH levels which causes the follicle to be unable to continue growth beyond this point. Lame cattle lacked an LH surge in res ponse to exogenously administered estradiol (Dobson and Alam, 1987). In another study, la me cattle were more likely to develop follicular cysts than non-lame cohorts (Mele ndez et al, 2003). A po ssible explanation for the underlying mechanism is through stress causi ng a disruption in the GnRH/ LH surge. The role of stress in the pathogenesis of ova rian cysts has been partially elucidated through studies examining the endocrinology of cows with ACTH-induced cysts. The hormonal profile in cows with cysts induced by estradiol compared to ACTH is different yet their response to subsequent estradiol ad ministration is simila r. Estradiol induced cysts occurred as a result of an estradiol induced LH surge in the absence of a preovulatory size follicle and prior to comp lete luteal regression, while ACTH induced cysts were the result of the absence of an LH surge (Refsal et al., 1987). Similar to spontaneously occurring and estradiol-induced cysts, cows with ACTH-induced follicular cysts did not have an LH surge in respons e to exogenously administered estradiol (Ribadu et al., 1999). This suggested that co ws with ACTH-induced cysts share a similar hypothalamo-pituitary lesion to cows with spontaneous ovarian cysts. A proposed mechanism for stress/ ACTH induction of cysts involves suppression of the LH/FSH surge and decreased follicul ar steroidogenesis. This is through the inhibitory effects of increased cortisol and progesterone produc tion by the adrenals. Progesterone can block the GnRH/ LH su rge at the hypothalamus while cortisol suppresses estradiol secretion and LH receptor content in granulosa cells (Kawate, 2004). The combination of these effects cause the fo llicle to produce an ineffective amount of
51 estrogen for a GnRH/LH surge, compounded by suppression of the GnRH surge at the hypothalamus, resulting in continued fo llicular growth and cyst formation. The source of elevated ACTH is another area under investig ation for its role in the pathogenesis of cysts. One hypothesis involve s a lack of cortis ol-mediated negative feedback causing increased ACTH levels. De creased circulating cortisol may be caused by the dysfunction of enzymes responsible for ox idizing cortisol to cortisone. Thurston et al., (2003) found that cysts contained more inhibitors of the Type 1 11 HSD, a ketosteroid reductase enzyme i nvolved in conversion of cortisone to cortisol, compared to follicular fluid from large antral follicles. These inhibitors could potentially also effect enzyme activity in the liver and other locations to cause an overall decrease in circulating cortisol, therefore altering negative f eedback of ACTH (Thurston et al., 2003). Another possible role of stre ss in the pathogenesis of cysts was examined in a study performing immunohistochemistry on pituitary cells of normal cyclic cows and cows with cysts (Busato et al., 1995). There wa s a deficiency in cell number and staining intensity of gonadotropes immunoreactive fo r LH and evidence of hyperactivity for corticotropes in cystic cows. It is, therefore, possible that th e increase in ACTH secretion and the subnormal activity of LH cells may be secondary to the act ivation of the ACTH cells, although it is possible that the same mechanism ma y alter both cell populations. Treatment Ovarian cysts which occur prior to the fi rst postpartum ovulation are more likely to undergo spontaneous recovery, with a resoluti on rate of 5060%, than those occurring after the 1st postpartum ovulation with a resoluti on rate of 20% (Youngquist, 1986; Peter, 1997). Morrow et al., (1966) reported a spontane ous recovery rate of 48% and a higher recovery rate among cysts dia gnosed prior to the first postp artum ovulations than those
52 developing after the first postpartum ovulat ion, although this was not statistically significant. In a study following 42 cows with cysts identified in the first 90 days post partum, 12 spontaneously recovered within 30 days of diagnosis, another 12 recovered between 31-90 days, 5 recovered between 91-168 and 13 had not recovered by 300 days (Whitmore et al., 1974). The overall recovery rate by 300 days in this study was 69%. Another study reported the mean cyst durati on following detection to be 31.0 4.3 days (Carroll et al., 1990). Of those cows under going spontaneous recovery, approximately 35-45% will have a repeat in cidence (Peter, 1997). Manual rupture of the cyst ha s been recorded with vari ous recovery rates. This treatment can cause the undesi rable effects of hemorrhage and adhesions on the ovary and ovarian bursa (Roberts, 1971; Younquist 1986). Manual ruptur e did not restore hypothalamic responsiveness to subsequent estr adiol administration in all of nine cows while pretreatment with progesterone restor ed responsiveness in all of seven cows (Nanda et al., 1991). This indicates that al though the cyst may not be physically present on the ovary, as in the case of manual rupture, a lesion still pers ists at the hypothalamus. Hormonal treatments for ovarian cysts have included human chorionic gonadotropin (hCG), gonadotropin releasi ng hormone (GnRH), progesterone and prostaglandin. Hormonal treatment is prefe rred over the uncertainty of spontaneous recovery and the damage caused by manual rupture. The general purpose of GnRH treatment is to induce an LH surge, causing fu rther luteinization of the cyst or ovulation of a responsive growing follicle within a deve loping wave. Luteinization or the presence of a CL will increase circulating progeste rone concentrations, decrease basal LH concentrations and potentially restore hypot halamic responsiveness to estradiol and
53 normal cyclicity. Treatment with hCG simula tes the LH surge and acts directly on the ovary to induce further luteini zation of the cyst. Administra tion of 10,000 units of hCG to 21 cows resulted in a 85.7% (18/ 21) recovery rate with cows returning to estrus within mean interval of 20.5 days (Morrow et al., 1966 ). Despite the effec tiveness of hCG as a treatment, and due to its large molecular si ze and immune interactions, it has the negative side-effects of potential severe reactions and a loss of effectiv eness over time. It is also a more expensive and less stable compound th an GnRH (Archbald and Thatcher, 1992). Administration of GnRH to cystic cows results in return to cyclicity in approximately 75-80 % of cases (Archbald a nd Thatcher, 1992; Fari n and Estill, 1993). The first study using GnRH to treat cysts i nvolved 5 cows treated with 100ug of GnRH IV, three times 120 minutes apart and found that all cows responded with an increase in LH (Kittok et al., 1973). The greatest increment in LH le vels occurred after the 2nd injection and estrus was observed in all cows within 20-24 days. Biershwal et al., (1975) evaluated the effectiveness of a single IM injection of GnRH at doses of 0, 50, 100 and 250ug. They found that 50, 100 and 250ug doses we re statistically equal although there was a slightly better response to the 100ug dose and all were better than 0ug (Bierschwal et al, 1975). In a companion paper, of 18 cows treated with 50-250ug of GnRH, 72% were observed in estrus within 20 1.5 da ys. Cows responding positively had an increase in LH and progesterone on day 0, and increased estrogen levels from days 1-13 as well as an increase in firmness of the cyst with a de crease in ovarian size (C antley et al., 1975). Whitmore et al., (1979) observed a 76% reco very rate in 225 cows treated with 100ug GnRH, with estrus occurring in 15-30 days. They also examined the effect of repeated
54 injections and found no decrease in effectiveness with subseq uent injections at 2 to 4 week intervals as may be e xpected with hCG treatment. A study examining the preventative use of GnRH for ovarian cysts compared the frequency of cysts and reasons for culling in a group of 204 cows treated with GnRH or saline at 14 days post partum (Britt et al., 1977). These authors found that fewer GnRH treated cows developed cysts and were culled for infertility reasons compared to saline treated cows. Although GnRH was determined to be an effective treatment of ovarian cysts, the interval from treatment to subsequent es trus detection was a pproximately 15-30 days. Induction of luteolysis using prostaglandin F2 (PGF2 ) could potentiall y shorten the time to estrus after GnRH caused luteinization of the cyst or another follicle. Kesler et al. (1978) compared the treatm ents of GnRH only, PGF2 only and GnRH followed by PGF2 in 9 days in cystic cows. The treatment of GnRH followed by PGF2 in 9 days resulted in the most consistent return to estrus while PGF2 only resulted in inconsistent and variable intervals of retu rn to estrus (Kesler et al ., 1978). Another study comparing GnRH followed by PGF2 in 14 days to GnRH alone and insemination at observed estrus in both groups found no difference in the c onception and pregnanc y rates of the two groups (Archbald et al., 1991). Since the succe ss of treatment in this study depended on expression of estrus, an important and limiti ng factor to the treatment success, when defined as conception or pregnancy, is the estrus detection rate. The use of protocols using GnRH and PGF2 to synchronize ovulation, allowing for timed insemination without the need for detection of estrus, have proven to be effective in lactating dairy cows (Pursley et al., 1995; Momcilovic et al., 1998). The
55 effectiveness of such protocol s in dairy cows with cysts has been evaluated and compared to other traditional methods. Normal cows s ubjected to a protocol using 100ug of GnRH on day 0, 25 mg PGF2 on day 7, 100ug GnRH on day 9 and timed insemination 16 hours later (Ovsynch) have typically achieve d conception rates of approximately 31-32% (Bartolome et al., 2000; Gumen et al., 2003). It was speculated that this protocol would work well in cystic cows as well as eliminate the need for estrus detection which has been an inhibitory factor in the success of some treatments. Bartolome et al. (2000) compared the Ovsynch protocol to GnRH on day 0 followed by PGF2 on day 7 and insemination at detected estrus as treatments for cystic ovarian disease. Although the conception rate of cows in the Ovsynch protocol was significantly lower than those inseminated at detected estrus (23.6% v. 51.7%), the overall pregnancy rate was not different between the two groups (23.6% vs. 18%). This study also f ound a significantly lower pregnancy rate among all the cystic cows compared to normal cows. In a similar study comparing two protocol s relying on insemination at detected estrus, one group received both GnRH and cl oprosternol on day 14 and were observed for estrus while a second group received Gn RH and cloprostenol on day 0, followed by cloprostenol on day 14 and then were obser ved for estrus (Lopez-Gatuis and LopezBejar, 2002). These authors found the sec ond group receiving the GnRH and PGF2 on day 0 and closprostenol on day 14 had a lowe r cyst persistence rate, better estrus detection rate and a higher ovul ation rate. In part two of th e same study, a protocol using Ovsynch with an additional cloprostenol injection on day 0 along with the GnRH on day 0 was compared to the standard Ovsynch prot ocol. The best result s were obtained from the protocol using cloprostenol and GnRH at the same time on day 0 for ovulation rate,
56 pregnancy rate and less cyst persistence. The major benefit of this protocol was that it allowed for detection of estrus in the week after the first cloprostenol injection and prior to the timed insemination. When Ovsynch and estrus detection were compared for a 21-day period in both ovular and anovular cows with varying follicular sizes, ther e was no significant difference in the conception rate for the two protocols (Gumen et al., 2003). Ovular cows had a significantly better con ception rate than anovular cows (32 vs 9%). Anovular cows in the estrus detection group had a 42% spont aneous recovery rate. Although not all the anovular cows in this study we re cystic, approximately 20% of those cows would be considered cystic by standard definitions a nd a further 58% had follicles greater than 15 mm persisting for two weeks prior to diagnosis. The regular use of bST (bovine somatotrop in) to increase milk production in dairy cows throughout their lactation has lead to investigations on its reproductive effects. In normal cows, bST resulted in a significantly lower conception rate when administered 46 days prior to initiati on of the Ovsynch protocol but not if administered 1-3 days prior to Ovsynch (Bartolome et al., 2002). In that study, the interval from the bST injection to initiation of the Ovsynch protocol had no effect on conception rate in cystic cows. In another study, pretreatment of cystic co ws with bST and GnRH, GnRH, bST or no treatment prior to Ovsynch found the best conception rate in the group receiving no pretreatment (Bartolome et al., 2003). Th e group receiving bST only had a significantly lower conception rate than the group receiving no pretreatment. Pretreatment with GnRH is thought to incr ease the chance of th e presence of luteal tissue or a CL at the time of the PGF2 injection and to synchronize follicular wave
57 growth. A comparison of Ovsynch and a Gn RH + Ovsynch protocol in cystic cows resulted in a significantly higher conception rate for the GnRH + Ovsynch group over the Ovsynch group (30.0% vs. 20.2%, re spectively; Bartolome et al., 2005a). This study also found there was a significant inte raction between type of cyst and treatment, where cows with a diagnosis of follicular cyst had a better pregnancy rate in the GnRH+ Ovsynch protocol over the Ovsynch protocol. There was no significant diffe rence in pregnancy rate for cows with luteal cysts for either protocol. Progesterone as a treatment for cystic ova rian disease has been used both for its ability to lower LH concentr ations and to reset the hypothalamus responsiveness to the positive feedback effects of estrogen. A shift in serum progesterone from low to high (0.5-5.0 ng/ml) concentrations caused LH pul se frequency and amplitude as well as estradiol concentration to decr ease significantly within 6 hours (Bergfeld et al., 1996). In cows treated with estradiol and progesterone together, the suppressive effect on LH was greater than with either ster oid alone (Stumpf et al., 1993). A significant regression in the size of the cyst occurred following treatment with progesterone when compared to cystic cows with no treatment during the same time period (Todorki et al., 2001; Calder et al., 1999). Some evidence for the role of proge sterone in restoring hypothalamic responsiveness to es tradiol came from a study by Nanda et al. (1991). These authors examined the results of estradiol treatment in cystic cows before and after administration of progesterone and found that only a propor tion responded to estradiol given before progesterone whereas all responded with a Gn RH/ LH surge after progesterone exposure for 7 days (Nanda et al. 1991). Similar results were found in a study in which ewes were
58 exposed to high levels of estradiol for 12 days and became unresponsive to further estradiol injections (Ozturk et al., 1998). In this study, ad ministration of progesterone restored the ability of the hypothalamus to res pond to estradiol. In two studies by Gumen et al. (2002), a large follicle anovulatory condi tion similar to cysts was induced in cows, and in every case these cows were not responsive to exogenous estradiol until treated with progesterone. Treatment with GnRH con tinued to produce an LH surge in all cows but estradiol did not, unless the cow had r eceived progesterone (Gumen and Wiltbank, 2002; Gumen et al., 2002). Gumen and Wiltbank (2005) performed two further experiments to define the effect of CIDR in serts on progesterone leve ls and the length of exposure needed to restore hypothalamic res ponsiveness to estradiol. In the first experiment, cows with induced cysts were tr eated with a CIDR (1.9g progesterone) for 0, 1, 3 and 7 days. The results indicated that th e CIDR caused serum progesterone to reach a steady state concentration of 1.3ng/ml within 3 hours, and the minimum exposure time required to restore hypothalamic responsivene ss to estradiol was 3 days. In the second experiment using anovulatory cows, some of which would be considered cystic, treatment with a CIDR (1.9g progesterone) for 0, 1 or 3 days indicated the best response was achieved in cows treated for 3 days. All of the cows treated for 3 days ovulated within 1 week after CIDR removal. The efficacy of progesterone as a treatmen t for spontaneous cysts has been studied alone or in combination with GnRH and/or PGF2 In a trial of cystic cows treated with 2 progesterone-releasing intrav aginal devices (PRID, 1.55g pr ogesterone each; total of 3.10g) for 9 days, treatment resulted in d ecreased circulating LH within 1 day and induced a new follicular wave (Calder et al., 1999). All PRID cows ovulated within 3 to 4
59 days after PRID removal. This treatment also resulted in a significant decrease in cyst size compared to untreated c ontrols. Another study used beef donor cows with ovarian cysts and treated them with a CIDR (1.9g pr ogesterone) for 14 days (Todorki et al., 2001). During the CIDR treatment, 2 or 3 fo llicular waves emerged and regressed, with an average of 7 days between each waves. Th e ovulatory follicle had an average period of 7.6 days from emergence to ovulation. Seru m progesterone levels increased from 0.6ng/ml on day 0 to 2.6ng/ml by day 1. Seru m progesterone then decreased from 2.6ng/ml on day 1 to 1.3ng/ml on day 14. All the CIDR-treated cows ovulated after CIDR removal and continued to display normal estr ous cycles for the next two months, while control cows did not. The fact that normal estrous cycles continued for a period of two months is a good indication that the underlying hypothalamic de fect was corrected by the treatment. In a recent study, 8 cows with cyst s, 3 of which also had a CL, were treated with a CIDR for 9 days, and also given GnRH on day 0 and PGF2 on day 7 to synchronize follicular wave development. Of these cows, 7 out of 8 ovulated following CIDR removal (Ambrose et al., 2004). Collect ively, these studies provide evidence that progesterone should be an e quivalent or better treatment than GnRH alone or the Ovsynch protocol for cows with cystic ovarian disease. Economics of Cystic Ovarian Disease Cystic ovarian disease represents a c onsiderable economic concern for dairy managers. The high lactational incidence rate of 7-25% make s this disease a major cause of reproductive insufficiency in dairy co ws (Peter, 1997). The rate of spontaneous recovery is low and variable, ranging fr om 20% (Youngquist, 1986; Peter, 1997) to 48% (Morrow et al., 1966) and with an average duration of 31.0 4.3 days (Carroll et al., 1990). The time from diagnosis to spontan eous recovery is unpredictable with
60 approximately 58% recovering within 90 days and 30% still unresolved by 300 days (Whitmore et al., 1974). Among cows undergoi ng spontaneous recovery, approximately 35-45% will have a repeat incide nce (Peter, 1997). Even with tr eatment, cystic lactations have a longer interval from calving to con ception when compared to non cystic herdmates (+33.5 daysBartlett et al., 1986; +64 daysBosberry and Dobson, 1989). Therefore it is generally more economical to treat the condition whenever it is identified. Furthermore, cows with ovarian cysts ha ve an increased lik elihood to be culled. The percentage of cystic cows culled during th e affected lactation, or the disease specific culling risk, was 20.9 % (Grhn et al., 1998). In a slaughter house survey, 85% of cows culled for infertility had cy sts (Kubar and Jalakas, 2002). In another study, cows with cystic lactations had a 1.23 relative risk for being culled than cows with non cystic lactations and the average cost attributable to culling alone was $43/ case of cystic ovarian disease (Bartlett et al., 1986). Previous reports analyzing the economic as pects of cystic ovarian disease have focused on the total economic impact of the condition per lactation (Bartlett et al, 1986) or used the technique of deci sion tree analysis to determine whether it is beneficial to treat cystic ovarian disease or wait for s pontaneous recovery (White and Erb, 1980) and at what day postpartum should cows be scre ened for the condition (White and Erb, 1982). In the study by Bartlett et al. (1986), costs considered for th e economic analysis included the average additional non-pre gnant days, additional cows culled, additional semen costs, additional reproductive exams, drugs and labor costs. Estimates for the cost of semen, labor, drugs, reproductive exams, replacement and salvage costs were obtained from a producer survey. The cost for an additional day open was determined using estimates
61 from the literature. The incidence rate, culling risk and average nonpregnant days for cows with cystic ovarian disease were obtaine d from available herd data. The total costs were added together to obtain the average co sts associated with cystic ovarian disease and the result was $137/ cystic lactati on. Although this study provides a rough estimate of the total cost associated with cystic ovari an disease, it does not take into account the future loss in production in subsequent la ctations or other probabilities, such as pregnancy loss and involuntary cu lling. It also does not cons ider that these cows would be subject to the herd culli ng policy and culled once they me t any of the culling criteria. It is also likely that conditions on a modern dairy farm are very different from the farms used in that study. Based on this analysis, it is obvious there is a considerable economic impact of cystic ovarian di sease. Therefore, future emphasis should focus on optimizing treatment decisions from both a clinical a nd economical aspect, since the condition can not be prevented. One method of determining the economic impact of a treatment choice is the decision tree technique (White and Erb, 1980) This method considers the chances of different outcomes due to a decision and multip lies the probability of an outcome by the total cost associated with it. These are summed for all the possible outcomes based on a decision and provide an estimate for the cost of that decision. When there are two possible decisions, the decision w ith the least associated costs is chosen and the other is excluded. Each possible decision is analyzed in this way progressing from the future to the present. Finally a value can be applie d to decision choices in the present that represents all the future probabilitie s and costs based on that decision.
62 In the early 1980s, two studies were publishe d using the technique of decision tree analysis for decisions involving cystic ova rian disease. The objective of the 1st study was to determine which day postpartum it would be cheaper to wait for spontaneous recovery than to treat cows with cystic ovar ian disease (White and Erb, 1980). The 2nd study was a follow up to the first and determined the id eal day postpartum to screen all cows for cystic ovarian disease (White and Erb, 1982). Using this technique, it was determined that it was always more advantageous to treat cows for cystic ovarian disease than to wait for spontaneous recovery (White and Erb, 1980). The only costs considered in this analysis were the additional days open, the additional reproductive exams and drug costs. In the follow up study, it was determined that th e ideal day postpartum to screen cows for cystic ovarian disease was at day 45 (W hite and Erb, 1982). These studies did not consider many factors in their analysis but utilized a sound approach for estimating the costs associated with a decision based on future probabilities. Computer models using dynamic programming to simulate dairy herds over time provide a realistic method of de termining the future profit of a cow (de Vries, 2004). This type of economic analysis takes into consid eration more factors and follows cows for a longer period of time, or for as long as a pr ofit is affected, and c onsiders the realistic probability that pregnancy may be lost at any time and the risk of involuntary culling. Dairy herd simulation models can be used to measure the economic impact of decisions or events such as changes in heifer replacement policy (d e Vries, 2004), conception rate and estrus detection efficiency (de Vries and Conlin, 2004). There have not been any reports comparing the economics of two differe nt treatment protocols for cystic ovarian disease using conditions repres entative of modern dairy fa rms. An economical analysis
63 using this type of dairy herd simulation m odel may prove to be an effective method for the comparison of two potential treatment choi ces for cows with cystic ovarian disease. Timed Insemination Programs in the Dairy Cow Timed insemination programs eliminate the need for estrus detection, decrease the number of days open beyond the voluntary wait ing period and decrease the variation in monthly pregnancy rates. Heat detection rate, or estrus detection efficiency, can be defined as the number of cows observed in estrus divided by the number of estruses which should have occurred in a 21 day period (Heersche and Nebel, 1994). Heat detection rate has a greater impact on the number of days to 1st service, number of days open, and the length of the calving interval than either the conception rate or reproductive culling policy (Rounsaville et al., 1979). Furthe rmore, an increase in the heat detection rate from 35%-55% decreases the average days to 1st service by 1 day (Rounsaville et al., 1979). The economic significance of a day ope n is that for each additional day open beyond the voluntary waiting period the tota l milk production is decreased by 2.4kg of milk and 0.112kg of fat (Louca and Legates, 1967). In another study, each day open beyond the voluntary waiting period decrease d the annual milk production by 4.5 kg in 1st lactation cows and by 8.6 kg in 2nd+ lactation cows (Olds et al., 1979). Furthermore, each additional day open between 40 and 140 DIM decreased the income over feed costs by $0.71 for 1st lactation cows and by $1.18 for 2nd+ lactation cows. Estrus detection is an ev en greater challenge in larg e dairies where labor may be limited and other factors limit estrus behavi or expressed by the cows. An average cow has an estrus period of 5.8 hours in which sh e displays an average of 6.7 mounts lasting for 3.2 seconds each (At-Taras and Spahr, 2001). This indicates that there is an average of 22 seconds over a 5.8 hour period where the cow could be observed in estrous. In hot
64 weather, total duration in which estrus behavi or is displayed decreases to an average of 2.9 hours (At-Taras and Spahr, 2001). The use of bST also tends to decrease estrus detection rates, particularly in multiparous cows (Santos et al., 2004b). When cows were observed for 30 minutes twice a day, greater than 50% of ovul ating cows did not display mounting behavior and instead displayed other secondary signs of estrus such as chin resting, restlessness or mucous vulvar disc harge (Van Eerdenburg et al., 2002). These cows which showed secondary behavioral si gns of estrus were more likely to ovulate greater than 24 hours from the time first ob served. Estrus detection accuracy, or the proportion of cows observed in estrus that are truly in estrus, is another potential source of error and cause of decreased conception rates. The error ra te for estrus detection is estimated to be greater than 20% in approximately 30% of dairy herds and a significant cause of decreased conception in th ese herds (Nebel et al., 1987). Programs using Prostaglandin F2 Alpha In 1972, Rowson et al. determined that intrauterine administration of PGF2 into the horn ipsilateral the CL between days 5 and 16 of the estrous cycle resulted in estrus by the 3rd morning following administration. These au thors also determined that the same treatment between days 1 and 4 of the estr ous cycle was ineffec tive. This study opened the door to the use of prostaglandin as a poten tial tool for controlli ng the estrous cycle of dairy cattle. It would therefor e be possible to synchronize co ws so that estrus detection could be condensed into a shortened period of time and decrease the labor required. When PGF2 was administered to lactating da iry cows between days 6 and 17 of the estrous cycle, 85% displayed estrus within 144 hours (Macmillan et al., 1978). If PGF2 is administered to random cows without knowledge of the st age of the estrous cycle, any cows which by chance are between days 1 and 5 will not display estrus.
65 Whereas any cows which are at day 17 or grea ter will naturally display estrus during the expected time period. It was then hypothesized that the administra tion of two doses of PGF2 at an interval of 11 to 14 days w ould improve estrus synchrony. The rationale behind this protocol was that cows which were unresponsive to the first injection because they were either in proestrus or metestrus at that time would be betw een days 6 and 17 of the estrous cycle by the time of the second inj ection, or 11-14 days later. This theory was tested by Cooper in 1974, who gave 2 doses of PGF2 11 days apart to heifers and observed that 90% displayed estrus within 48-72 hours after the 2nd injection and another 6% displayed estrus between 72 and 96 hours. Jackson et al. (1979) also gave two injections of PGF2 11 days apart and found that estrus occurred earlier and in a more narrow range than with only 1 injection. Another study determined that luteolysis occurred after in the 1st injection in 60% of cows and after the 2nd injection in 72% of cows (Stevenson et al., 1987). The use of two doses of PGF2 at an interval of 14 days is widely applied in modern dairy herd management as a means of setting up cows for other timed insemination programs and as prevention for endometritis and pyometra. Pankowski et al. (1995) administered sequential injections of PGF2 14 days apart and compared the degree of synchronization and resulti ng fertility for 3 sequential injections of PGF2 2 sequential injections of PGF2 and traditional methods of managing postpartum cows involving palpation per rectum and intr auterine infusion. In that st udy, the protocol involving three sequential injections of PGF2 was the most cost effective method and resulted in the greatest synchrony of estrus following the thir d injection. Despite th ese advantages, there was no significant effect of treatment on fertil ity. Lopez-Gatius et al. (2003) administered
66 2 doses of PGF2 at days 22 and 36 postpartum then subjected cows to a standard reproductive examination on days 50 and 71 postpartum. In that study, cows in the treated groups had fewer cases of cystic ovarian disease and pyometra, an increased proportion with luteal activity, a higher ovulation rate, a greater number in estrus, and a higher pregnancy rate. The interval between PGF2 administration and estrus is highly variable among cows when it is given at random stages of th e estrous cycle. The protocol using 2 doses 11-14 days apart decreases the variability in the interval between administration and estrus, therefore increasing the synchrony. The time period between luteolysis and estrus is thought to be related to the maturity of the follicle at the time of PGF2 administration. If the follicle is large, then time to estrus is decreased whereas if it is small, more time is required for the follicle to reach ovulatory size. The underlying follicular waves play a significant role in determining the time to estrus when PGF2 is administered in the early to mid-luteal phase. Jackson et al. (1979) recogn ized that the time to estrus was shorter in cows treated with PGF2 on days 7-8 and days 15-16 than on days 12-14. These authors also suggested that the presence of a large fo llicle was the factor involved in shortening the time to estrus and de creasing the variability. Similar to the above findings, when PGF2 was administered early in the estrous cycle (days 5-9) the interval to estrus was s horter than when it was administered later in the estrous cycle (King et al., 1982 ). When heifers were given PGF2 on day 7 or 15 of the estrous cycle, 88% and 73% respectively, were in es trus between 32 and 56 hours post PGF2 ; whereas when it was given on day 11, only 13% were in estrus between 32 and 56 hours post PGF2 (Tanabe and Hann, 1984). Later us ing ultrasound, Savio et al.
67 (1990) determined that the 1st wave dominant follicle of the estrous cycle will ovulate following PGF2 administration on day 7. That study confirmed, through the use of ultrasound, the hypothesis that the 1st wave dominant follicle has ovulatory capacity if luteolysis occurs and that the presence of a dominant follicle increases the degree of estrus synchronization following PGF2 Another interesting findi ng was that the interval from PGF2 to estrus was generally shorter in he ifers than in cows (King et al., 1982). Although PGF2 was useful for synchronization of estrus, as long as estrus detection was still required it continued to be a limiting factor in the overall reproductive efficiency. Programs where insemination could be performed at a predetermined time and completely eliminate the need for estrus dete ction would be ideal, but only if acceptable fertility could be achieved. In an attempt to develop this type of program, Lauderdale et al. (1974) compared three groups of cows: 1) cows observed for estrus and inseminated, 2) cows treated with PGF2 and inseminated at detected estrus, and 3) cows treated with PGF2 and timed inseminated at 72 and 96 hours. In that study, conception rates (52.255.8%) and overall pregnancy rates (30-42%) were not significantly different between treatment groups which indicated that acceptabl e fertility could be achieved using timed insemination. Macmillan (1978) examined a protocol where cows were given PGF2 between days 6 and 17 of the estrous cycle and were timed inseminated 72 hours after PGF2 and cows which were observed in estrus after th e first insemination received a second timed insemination at 96 hours post PGF2 This protocol resulted in a 57% conception rate but did not entirely eliminate the need fo r estrus detection. When 2 doses PGF2 were given 11 days apart and insemination was timed at 72 and 96 hours post PGF2 the conception
68 rate was 39% in heifers (Macmillan et al ., 1978) and 30% in l actating dairy cows (Stevenson et al., 1987). A single tim ed insemination at 80 hours post PGF2 resulted in a 23% (Stevenson et al., 1987) to 46.2% (King et al., 1982) conception rate in lactating dairy cows and a 46.7% conception rate in heifers (King et al., 1982). When a single timed insemination was used at 80 hours post PGF2 the stage of the estrous cycle at which PGF2 was given significantly influenced the conception rate in heifers. Heifers given PGF2 early in the estrous cycle (days 5-9) had a significantly decreased conception rate, likely due to the fact that the majority were in estrus 30 hour s prior to insemination (King et al., 1982). There are a proportion of cows with matu re corpora lutea which do not respond to an intramuscular injection of PGF2 It is possible that in these cases, repeated administration of PGF2 may be needed to achieve lute olysis. Archbald et al. (1993a) tested protocols administering two doses of PGF2 8 hours apart and 24 hours apart and compared these to the standard single dose. The results of that study indicated that 2 doses of PGF2 8 hours apart increased th e number of cows in estrus compared to the other two protocols and di d not affect fertility. Programs using Prostaglandin F2 Alpha and GonadotropinReleasing Hormone In 1974, Kaltenbach et al. determined that GnRH given intramuscularly resulted in an LH and FSH surge, with peak levels o ccurring within 100 minutes after the injection. In that study, estrus and ovulation occurred in most cows within 24 hours. This opened up possibilities for the use of GnRH in the synchronization ovulation following an injection of PGF2 In the same study, Kaltenbach et al. (1974) also attempted to synchronize ovulation after PGF injection but failed due to failure of the PGF2 in causing luteolysis. Fernandez-Li ma et al. (1977) gave PGF2 followed by treatment with
69 a GnRH agonist (D-ala-GnRH) 64 hours later, this treatment either induced an LH surge or potentiated a natural LH surge. These resu lts agree with those of Kaltenbach et al. (1974) and indicate that GnRH has the potenti al to improve synchronization of ovulation when used in combination with PGF2 The next challenge was to determine th e exact combination and timing of GnRH and PGF2 administration that would produced an acceptable conception rate, eliminate the need for estrus detection and decrease and reduce the variability around the calving interval. Rodriguez et al. (1975) compared three different prot ocols; a control group with standard estrus detecti on, a group receiving PGF2 followed by estrus detection, and a group receiving PGF2 + GnRH in 48 hours + timed insemination 15 hours later. The timed insemination protocol using GnRH resulte d in a conception/ pregnancy rate of 22% which was significantly lower than that obs erved in the control group (36%). In a comparison of three different protocol s using sequential in jections of PGF2 and timed insemination with or without GnRH to a standa rd protocol of estrus detection, conception rates were lower in the timed inseminati on groups and the calving interval was not reduced, although the variability in the interval to 1st service was significantly reduced (Lucy et al., 1986). In a study where GnRH was given at the time of insemination and 72-80 hours after PGF2 GnRH treatment improved fertility compared to no treatment or PGF2 given concurrently with timed insemination (Archba ld et al., 1992). In that study, the best pregnancy rate was still obtained with insemi nation at detected estr us, but there was an effect of season where cows treated with GnRH in the spring but not in the summer had higher conception rates. When GnRH was given at the time of estrus and insemination to
70 repeat breeder cows in an attempt to ensure ovulation and improve c onception rates, there was no significant effect on fertility when co mpared to no treatment (Archbald et al., 1993b). Administration of GnRH causes ovulation of a dominant follicle, if present, and removes the estradiol inhibition on the anterior pituitary resulting in an FSH surge and the emergency of a new follicular wave. It was therefore speculated that GnRH could also be used to control follicular wave deve lopment and ensure that a dominant follicle would be present on the ovary at the time PGF2 treatment. Furthermore, it was speculated that treatment with GnRH 6 or 7 days prior to PGF2 would also decrease the interval between PGF2 and estrus and increase estrus synchrony. In a study where cows were inseminated at detected estrus following PGF2 GnRH pretreatment 6 days prior to PGF2 increased the synchrony of estrus from 50% to 83.3% and decreased the variability in time to estrus while having no effect on fertility or the average time to estrus (Twagiramungu et al., 1992). A similar study in which cows were treated with either PGF2 or GnRH followed by PGF2 in 7 days and monitored for estrus, cows in the GnRH pretreatment group had a decreased in terval to estrus and improved synchrony (Wolfenson et al., 1994). Results from these tw o studies indicate that GnRH administered 6 or 7 days prior to PGF2 can alter follicle development and improve the synchrony following PGF2 A protocol (Ovsynch) was then devel oped in which cows were given GnRH, PGF2 7 days later, followed by GnRH in 48 hours. The first GnRH would either cause ovulation of a dominant follicle and emergence of a new follicular wave or coincide with the emergence of a natural follicular wave (Pursley et al., 1995). The PGF2 would cause
71 regression of the corpus luteum and the 2nd GnRH would cause ovulation of the newly developed follicle, usually within 24-32 hour s. Using this protocol, ovulation could be synchronized to an 8 hour period (Pursley et al., 1995). Furthermore, this protocol resulted in acceptable conception rates ( 50%Pursley et al., 1995; 33%Momcilovic et al., 1998; 35.3%Stevenson et al., 1996). Many researchers have compared the Ovsync h protocol to othe r accepted methods of estrus synchronization and timed insemi nation. Often the con ception rate to the Ovsynch protocol was less than observed with insemination at detected estrus (Stevenson et al., 1996; Stevenson et al., 1999). Despite a lower conception ra te, pregnancy rates with the Ovsynch protocol were frequen tly better than pregnancy rates following insemination at detected estrus (Steve nson et al., 1996; Momc ilovic et al., 1998; Stevenson et al., 1999). This is because there is no effect of heat de tection in the Ovsynch protocol as all cows are inseminated. Others have found similar conception and pregnancy rates between cows receiving th e Ovsynch protocol or inseminated after detected estrus, either natural or induced by PGF2 (Momcilovic et al., 1998; Burke et al., 1996; Pursley et al., 1997). Other major benefits to the Ovsynch pr otocol, besides potentially increasing the pregnancy rate, is a significantly decreas ed interval from calving to conception (Momcilovic et al., 1998), d ecreased median days to 1st service and more cows pregnant by 60 and 100 days post partum (Pursley et al., 1997). In disagreement with these findings, Burke et al. (1996) found there wa s no significant effect of Ovsynch on the interval from calving to conception or the ove rall pregnancy rate by 120 days postpartum, although they noted that the monthly con ception and pregnancy rates were more
72 consistent in cows receiving the Ovsynch protocol than in cows inseminated at detected estrus. Other research has focused on determining the ideal stage of the estrous cycle to initiate the Ovsynch protocol. Previous obs ervations have suggest ed that the Ovsynch protocol does not work as well if the 1st GnRH fails to cause ovulation of a dominant follicle, as may occur when a the largest follicle has not yet acquired ovulatory capacity. In general, follicles acquire ovulatory capaci ty just after deviation or when they are greater than 10 mm in diameter (Sartori et al., 2001). If the follicle is at a point just prior to deviation at the time of the 1st GnRH, luteal regression and ovulation may occur prior to the 2nd GnRH or the follicle will be undergoing atresia at the time of the 2nd GnRH. Vasconcelos et al. (1999) determined that the best response to the 1st GnRH injection was obtained when cows were between days 5-9 of the estrous cycle and the most significant factor influencing response to the 2nd GnRH was the response to the 1st GnRH. Furthermore, initiation of the Ovsynch protocol in the 1st half (days 1-12) of the estrous cycle resulted in a better sync hronization rate than when the protocol was initiated in the 2nd half of the estrous cycle (91% vs. 80%; Vasconcelos et al., 1999). In heifers, the ideal stage to initiate the Ovsynch protocol is betw een days 5 and 10 of the estrous cycle or in the early luteal phase, with the worst re sults occurring when protocol was started beginning on days 2, 15 or 18 (Moreira et al., 2000b). The amount of flexibility that ma y exist for the timing of the 2nd GnRH and the timed insemination has been another area of rese arch. It would be a major benefit to herd managers if acceptable fertility could be achieved without as much cow handling as required in the original Ovs ynch protocol. Pursley et al. (1998) found that acceptable
73 fertility was achieved if the timed insemi nation occurred anywhere between 0 and 24 hours after the 2nd GnRH, but not at 32 hours after. Furt hermore, the group inseminated at the same time as the 2nd GnRH (0h) had decreased pre gnancy losses and an increased female to male ratio. The flexibility in the timing of the 2nd GnRH in relation to the PGF2 injection has also been studied. Peters and Pursley (2003) f ound that fertility and follicle size at ovulation in creased linearly with increasing time from the PGF2 when they compared different protocols where the 2nd GnRH was given at 0, 12, 24, or 36 hours after the PGF2 There was also a tendency for an increased incidence of short luteal phases when the 2nd GnRH was given at 0 hours comp ared to at 36 hours. This study showed that although ovulation was stil l synchronized, fertility was compromised when the GnRH was given too soon after the PGF2 and that this may possibly be due to a shortened luteal phase. In partial agreement with this finding, Moreira et al. (2000b) found that decreased follicle size at ovulation was associ ated with decreased serum progesterone during the luteal pha se. It was also determined that the fertility of cows observed in estrus between the PGF2 and 2nd GnRH injection was normal if they were inseminated at that time (Kasimanickam et al., 2004). Body condition score has been found to signifi cantly affect conception rates to the Ovsynch protocol. Stevenson et al. (1999) found that for each 1 unit increase in BCS, conception rate increased by 10%, and co ws with higher BCS also had higher progesterone concentrations. Burke et al (1996) also found a positive correlation between BCS and pregnancy/ conception ra tes as well as between progesterone and conception/ pregnancy rates. When cows we re separated in two groups based on their BCS, cows with a low BCS (<2.5) had a signif icantly decreased conception rate to the 1st
74 service using the Ovsynch protocol than cows with higher BCS (Moreira et al., 2000a). In the same study, an economical analysis determ ined that if the percentage of low BCS cows in a herd was decreased from 30% to 10%, the economic benefit would be $10.33/ cow, or $10 330 in a 1000 cow herd. The economic benefit of applying the Ovsynch protocol to the standard reproductive management of an entire herd is another important area of concern. The Ovsynch protocol has a higher insemination submission rate than protocols relying on estrus detection and therefore has more cost associated with semen. There are also more injections required and more labor invol ved in cow handling. On the other hand, programs relying on estrus detection require more labor for the observation of cows. Ovsynch can potentially increase the overall pregnancy rate, decrease the interval from calving to conception, and decrease proporti on of cows open for long periods. These benefits may be limited in herds with excel lent estrus detection efficiency. In an economic comparison between the Ovsynch pr otocol and a protocol using PGF with insemination at detected estrus the Ovsynch protocol resulte d in a better pregnancy rate and was $29.44/ pregnancy cheaper (Britt and Gaska, 1998). In another study comparing insemination at detected estrus and the Ov synch protocol in 2 herds with different reproductive management, the Ovsynch protocol was better in the herd with lower estrus detection efficiency while the estrus detection protocol was better in the herd with better estrus detection efficiency (Te nhagen et al., 2004). Programs using Progesterone Progesterone is naturally rele ased by the corpus luteum in a normal estrous cycle and while progesterone is high, ovulation does not occur even though a dominant follicle may be present on the ovary. Exogenous pr ogesterone causes suppression of both
75 ovulation and estrus (Christian and Casida, 1948). Progesterone can be administered by daily feeding of oral progestins or subcutan eous progesterone impl ants. Stainless steel coils coated with silastic rubber impregnate d with progesterone and inserted into the vaginal results in adequate and more consistent plasma P4 levels than the method of daily feeding (Roche, 1976). Progesterone treatment has been used together with PGF2 in different protocols for the synchronization of ovulation and estrus. When a PRID was inserted for 7 days and PGF2 was given on day 6 of PRID treatment, or 1 day before removal (PRID7-PGF6), estrus synchronization was improved compared to untreated controls and protocol using a PRID for 6 days with PGF2 at the time of removal (Sm ith et al., 1984). In the same study, conception rate to timed insemination was better in cows treated with PRID7PGF6 over cows treated with two injections of PGF2 14 days apart. A comparison of two treatment protocols, both i nvolving two injections of PGF2 14 days apart but with or without insertion of a PRID for the final 7 days, found that the insertion of a PRID improved conception rates especially in cows with low progesterone at the beginning of the protocol (Folman et al., 1990). Similarly, a ddition of progesterone for the final 5 days between 2 injections of PGF2 14 days apart improved conception rates when compared to cows not receiving progesterone but not co mpared to untreated controls (Xu et al., 1997). Turnover of the dominant follicle is caused by the negative feedback effect of progesterone on LH and the removal of progest erone allows LH to rise and ovulation to occur. It has been discovered that when th e first wave dominant follicle develops in a milieu of low progesterone (2ng/ml), LH c oncentration is higher and the first wave
76 dominant follicle persists for many days than ovulates quick ly upon removal of progesterone (Savio et al., 1993). Conversel y, high progesterone concentrations (35ng/ml) suppress LH and cause turnover of the first wave dominant follicle and the emergence of a new follicular wave (Savio et al., 1993). Further evidence which supports this theory comes from a study comparing cows treated with either high or low progesterone for 10 days (Wehrman et al., 1993). Cows in the low progesterone group had increasing concentrations of estradiol and a shor ter interval to estrus upon removal of progesterone while cows in the high progest erone group had decreasi ng concentrations of estradiol, a longer interv al to estrus, and an improved conception rate. Elevated progesterone, either in the form of an additional injection or as endogenous progesterone from a CL, can also al ter follicular development. An injection of both estradiol and progesterone at the in sertion of a progesterone -releasing device in heifers, improved estrus response rates and increased synchrony (R oche et al., 1974). The underlying mechanism for this was partially eluc idated in a trial where heifers were fed melengestrol acetate (MGA) for 11 da ys, with an injection of PGF2 on day 2 of treatment and with or without an injection of progesterone on day 9 (Anderson and Day, 1994). The additional progesterone injection on day 9 caused atresia of the first wave dominant follicle and the emergence of a new follicular wave. Heifers in the progesterone group also had a higher conception rate while those not receiving the additional progesterone, and without a CL, developed persis tent follicles. Furthe rmore, the type of progesterone administered is important when no CL is present. In the absence of a CL, protocols using either a PRID or norgest omet both resulted in the development of persistent follicles, yet cows in the PRID group had a significantly higher conception rate
77 (Smith and Stevenson, 1995). These results em phasize the importance of both the type of progesterone and the presence of a CL. Progesterone protocols frequently also i nvolve the simultaneous use of estrogens. In cows given Syncro-Mate-B implants (ear implant containing norgestomet), estrodiol valerate (estra-1,3,5(10)-triene-3, 17-diol(17 )-, 17-pentanoate) causes regression of the CL and atresia of the 1st and 2nd largest follicles with the s ubsequent emergence of a new follicular wave (Bo et al., 1991). A protocol us ing an oral progestin treatment for 9 days with an injection of estadiol valerate on day 2 resulted in effective estrus synchronization and a similar conception rate to untreated controls (Wiltbank and Kasson, 1968). In the conditions of New Zealand, administration of a CIDR with an estradiol benzoate capsule for 8 days and PGF on day 7 resulted in 89% of cows exhibiting estrus within 5 days, but a lower overall conception rate than untreated controls (Xu et al., 1996). Follicle wave emergence can also be controlled by the addition of GnRH to a progesterone protocol. Through ovul ation of a responsive follicle, GnRH can be used to synchronize follicular wave development prior to progesterone removal or to synchronize ovulation after progesterone withdrawal. In a comparison of 3 groups given either GnRH, estradiol benzoate, or no treatment coinci dent with CIDR insertion, GnRH treatment caused progesterone to rise rather than decr ease as it did in the other two groups (Kim et al., 2005). In the same study, the GnRH group also had earlier follicular wave emergence, a larger ovulatory follicle and an increased pregnancy rate. Admini stration of GnRH at the same time as CIDR insertion cause d ovulation of a domi nant follicle and subsequently increasing progesterone concen trations throughout the treatment period (Ando et al., 2005). In that study, estrus synchronization wa s also improved in cows
78 receiving GnRH compared to cows not re ceiving GnRH. Administ ration of GnRH 30 hours after the removal of a progester one-releasing device improved both the synchronization of ovulation compared to c ontrols (Roche, 1975; doValle et al., 1997) and the final calving rate after a timed insemination (doValle et al., 1997). When GnRH is used to synchronize follic ular wave emergence at the beginning of progesterone treatment and to synchronize ovu lation after progesteron e withdrawal, with PGF2 given on day 7 to cause luteolysis; the protocol becomes synonymous to the Ovsynch protocol just with the addition of progesterone. Comparisons of Ovsynch to an Ovsynch + CIDR protocol indi cate that the addition of the CIDR significantly increases serum progesterone concentration over the Ov synch protocol alone (Kawate et al., 2004; El-Zarkount et al., 2004). The Ovsynch + CIDR protocol also improved conception rate compared to Ovsynch alone (72% vs. 47% ; Kawate et al., 2004). Another study found that the Ovsynch +CIDR prot ocol significantly improved c onception rate but only when the majority of cows were anestrous and the positive effect was lost when the majority of cows were diestrous (El-Zarkouny et al., 2004). In some regions where dairy cattle ar e grazed, calving is seasonal and estrus detection efficiency is good; the ideal estrus synchroniza tion protocol would result in tight synchronization of ovulation and estrus with good fertility, but does not necessarily require timed insemination. Studies in Irela nd have examined many different protocols for their ability to synchroni ze estrus and improve overall fertility. In one study, the protocol with the highest estr us detection rate and pregnanc y rate was one in which cows received a GnRH injection on th e day of CIDR insertion, PGF2 7 days later and removal of the CIDR 8 days later (GnRH+CIDR8-PGF7 ), this protocol was better than PGF2 +/-
79 eatradiol benzoate, and CIDR +/estradiol benzoate on the day of CIDR insertion (Ryan et al., 1995). In a subsequent study, the interval from CIDR withdrawal to estrus was shortened and the synchronization of estrus was improved when cows were treated with the same protocol (GnRH+CIDR8-PGF7) but w ith an injection of estradiol benzoate 10 hours after CIDR removal (Ryan et al., 1999). Another study in Australia compared two protocols, both using two in jections of estradiol benzoa te 9 days apart, with PGF2 on day 7, but with the insertion of an intravaginal progesterone releasing de vice (IVD) for either 7 or 8 days (Cavalieri et al., 2004). The prot ocol where the IVD remained in place for 8 days decreased the interval to estrus and improved synchronization compared to the 7 day protocol with no significant effect on fertility. Thes e results indicate that the synchronization of estrus through the use of progesterone can be improved by controlling follicular wave emergence and ovulation, using either GnRH or estrogens, in conditions where breeding is seasonal a nd estrus detection efficien cy is good. Furthermore, a protocol where the cow is exposed to proge sterone for 8 days, instead of 7 days, may improve synchronization.
80 MATERIALS AND METHODS This study was conducted in a large dair y herd (approximately 1,500 milking cows) in north east Florida (Baldwin, FL). These co ws were milked 3 times per day and had a rolling herd average 305-day milk producti on of 9000 kg. They were kept in open barns with dry manure bedding between milking. Between 60-63 days post partum, all cows were given bovine somatotropin every 14 da ys (bST; 500 mg, intradermally; Posilac; Monsanto Co, St. Louis, MO). Cows were fed a total mixed ration formulated to meet or exceed the requirements of the National Research Council (2001). The period of study was approximately one year, from October 13, 2003 to September 20, 2004. During this time, the farm was visited once a week. This dairy herd is on a reproductive herd health program admi nistered by veterinarians of the Veterinary Medical Teaching Hospital (Colle ge of Veterinary Medicine, University of Florida), and all reproductive, health, and management records were computerized. Cows were routinely vaccinated against bovine vira l diarrhea (BVD), infectious bovine rhinotracheitis (IBR), bovine respiratory sync tial virus (BRSV), parainfluenza virus (PI3), leptospirosis, campylobacteriosis, clostr idiosis, and gram-negative organisms (J-5 vaccine) according to recommendati ons by attending veterinarians. Using a protocol for synchronization of ovulation and timed insemination in this dairy herd, the conception rate of cows wit hout ovarian cysts was 30% while that of cows with ovarian cysts was 17%. It was anticipat ed that a protocol involving the use of a controlled internal drug releasing (CIDR) device with 1.38g progest erone will increase
81 the conception rate of cows w ith ovarian cysts to 30%. Therefore, a sample size of 130 cows per group was needed to demonstrate th at this difference will be significantly different at a level of P < 0.05 (95% confidence intervals and 80% power). A total of 401 lactating dairy cows with ova rian cysts were used in this study. This number was used to allow for the loss of co ws from the herd for various reasons. On the day of each visit (Day 0), cows diagnosed with ovarian cysts were al ternatively allocated to the 2 groups. The diagnosis of ovarian cy sts was based on per rectum palpation of the ovaries and uterus, and by ultr asonographic examination of th e ovaries. The criteria used on rectal palpation were the presence of mu ltiple follicles on the ovary with at least one follicle being 17 mm diameter (Halter et al., 2003) the absence of a corpus luteum (CL) on either ovary, and the lack of tonicity of the uterus (Archbald et al.,1991; Bartolome et al., 2002; Bartolome et al ., 2003; Zemjanis, 1970). On ultrasonographic examination of the ovaries, ovarian cysts we re recognized by the hypo echogenicity of the structure (Pierson and Ginther, 1984), and the absence of a CL on either ovary. Cows in the Ovsynch group (n = 201) were treated with GnRH (100 g, im; Cystorelin; Merial Limited, Is elin, NJ, USA) on Day 0, PGF2 (25 mg, im; Lutalyse Sterile Solution; Pfizer Animal Health, New York, NY, USA) on Day 7, GnRH (100 g, im; Cystorelin; Merial Limited, Iselin, NJ, USA) on Day 9, and timed inseminated 1620 h later, without detection of estrus. Cows in the CIDR group (n = 200) were treated with a CIDR (1.38g progesterone, intravag inal insert; EAZIBREED CIDR Pfizer Animal Health, New York, NY, USA) on Da y 0 for 7 days. On Day 7, the CIDR was removed, and cows were treated with PGF2 (25 mg, im; Lutalyse Sterile Solution; Pfizer Animal Health, New York, NY, USA). All cows in the CIDR group were observed
82 for estrus at least twice dail y, and cows exhibiting estrus following removal of the CIDR were inseminated according to the am-pm rule. Therefore, cows observed in estrus in the morning were inseminated in the afternoon, and cows observed in estrus in the afternoon were inseminated late that night. In either instance, cows were bred within 8-12 hours after the first sign of estrus. On Day 0, baseline data for parity (1st lactation= primiparous, 2nd + lactation= multiparous), days in milk (DIM), milk production (kgs/day) on the day of diagnosis, time of year of cyst diagnosis and body condition score (BCS; scale of 1 to 5; Ferguson et al., 1994) were recorded. Time of year was recorded to include the possible seasonal effects on reproduction in the analysis. Based on previous research (A l-Katanani et al., 1999), the cooler and more favorable time of ye ar for pregnancy is October to February, while the warmer and less favorable time of y ear is March to September. Therefore, cows enrolled from October to February and Marc h to September were recorded as receiving treatment in the cool and warm seasons, respectively. Pregnancy was determined in all cows between Days 30-31 after insemination using ultrasonography, and reconfirmed using pe r rectum palpation of the uterus and previously described techniques (Z emjanis, 1970) between 42-45 days. On Day 21, all cows were subjected to ovarian ultrasonography and per rectum palpation to determine the presence of a CL. Th e presence of a CL at this time was used to indicate that ovulation was induced, and that the protocol s resulted in resumption of normal ovarian cyclicity. Since production of progesterone by the CL is best determined by analysis of the peripheral plasma of the cow (MacDonal d, 1980), a blood sample for progesterone
83 determination was obtained on Day 21 to verify the presence/absence of a functional CL. Blood was obtained from the coccygeal vein into evacuated tubes (Vacutainer; BD, Franklin Lakes, NJ, USA) and immediately pl aced on ice. Samples were centrifuged at 3000 x g for 30 min, and serum was stored at -20 C until assayed for progesterone. Serum progesterone concentrations were determined using a sequential competitive chemiluminescent enzyme immunoassay (DPC Immulite Progesterone; Diagnostic Products Corporation, Los Ange les, CA, USA). A serum prog esterone concentration > 1 ng/ml was used to verify the presence of a functional CL and to determine the sensitivity, specificity, positive predictive value and negative predictive value of palpation per rectum for the diagnosis of the presence of a CL. Baseline data for parity, DIM, time of y ear, BCS and milk production on the day of diagnosis were compared using Chi-squa re and ANOVA (P < 0.05) respectively. Least squares means and ANOVA were used to determ ine the variables significantly associated with Day 21 serum progesterone concentra tion. The outcomes of interest for this experiment were the likelihood to be in seminated, presence of a CL on Day 21, conception and pregnancy rates for cows in each group. Conception rate was defined as the number of cows diagnosed pregnant divided by the total number inseminated. Pregnancy rate was defined as the number of cows diagnosed pr egnant divided by the original number enrolled in each group. Data for these variables were analyzed using logistic regression adjusting fo r parity, DIM, BCS, milk prod uction at diagnosis, and time of year. The explanatory variables were evaluated using the backward elimination procedure and variables that significantly a ffected the outcome remained in the model (Agresti, 1996). Treatment effect was forced to remain in the models. All possible
84 interactions between treatmen t and explanatory variables were tested before the final model was chosen. Statistic al significance was declar ed if P < 0.05 or the 95% confidence interval (95% CI) for the odds ra tio did not include 1. A 95% CI which does not include 1 indicates with 95% certainty that the true odds ratio is within the range of the confidence interval and repr esents a true association.
85 RESULTS Baseline Data Comparisons A total of 401 cows was enrolled in th e study of which 201 cows were enrolled in the Ovsynch group and 200 cows were enrolled in the CIDR group. Baseline data recorded included cyst diagnosis time of year (warm/cool), days in milk (DIM), parity [primiparous (PP)/ multiparous (MP)], body condition score (BCS), and milk production on the day of diagnosis (kg/day). The medi an DIM were 132 days. The median parity was 1.0, the median body conditi on score was 2.75, and the median time of year for cyst diagnosis was the warm time of year (March to September). The mean milk production was 29.5 kg/ day. Baseline data for cows in all groups are presented in Table 3. A Pvalue was determined based on ANOVA for continuous variables or 2 statistic for discrete variables. Baseline data for body condition score, milk production, parity, DIM, and season of cyst diagnosis were not signi ficantly different between treatment groups. Table 3. Baseline data comparisons for co ws in the Ovsynch and CIDR groups. Baseline Variable Ovsynch n=201 CIDR n=200 P Value Average BCS Range 2.75 2.0-3.75 2.75 1.75-4.0 0.7977 Milk Production, kg Range 29.5 9.9-52.2 28.3 12.6-45.0 0.3975 Parity% Primiparous 1st lactation / total 59% 119/201 61% 121/200 0.8538 Average DIM Range 158 61-510 169 59-521 0.2821 Season of diagnosis, % Cool # cool/ total 42% 85/201 40% 79/200 0.6275
86 Prior to further analysis, the contin uous variables of DIM and BCS were dichotomized into binomial variables based on median values, while milk production on the day of diagnosis was divided in 4 categ ories based on quartiles. The division into binomial and categorical variables was done to prevent loss of power. Days in milk were categorized into early and late lactation (< 132 days and >132 days, respectively). Body condition score was categorized into low and high based on a score of < 2.75 or > 2.75. There was an even distribution for the range of milk production on the day of diagnosis, which is presented in Figure 1. The four cat egories of milk production on the day of diagnosis were as follows; cows in Cate gory 1 producing <22.5 kg, cows in Category 2 producing > 22.5 and <28.8 kg, cows in Category 3 producing > 28.8 and <35.6 kg, and the top producers in Category 4 which were producing > 35.6 kg on the day of diagnosis. 0.0 10.0 20.0 30.0 40.0 50.0 60.0 050100150200250 CowsMilk Yeild (kg) Figure 1. The distributio n of milk production on the day of diagnosis (Kg) for cows in both groups combined.
87 Likelihood to be Inseminated, Return to Cyclicity, Conception and Pregnancy Rates Overall, the percentage of cows bred in the Ovsynch and CIDR groups was 82% and 44%, respectively. The percentage of cows returning to cyclicity in Ovsynch and CIDR was 83% and 79%, respectively. The c onception rate for cows in Ovsynch and CIDR were 18% and 23%, respectively. Pre gnancy rate for cows in Ovsynch and CIDR groups were 14% and 9.5%, respectively. All of these results are pres ented in Table 4 and Figure 2. Table 4. Summary of heat detection rate, conception rate, pregnancy rate and missing values/ lost cows by group. Items Ovsynch n=201 CIDR n= 200 % Inseminated 82% (164/ 201) 44% (87/200) % Return to cyclicity 83% (136/164) 79% (137/174) Conception rate 18% (29/158) 23% (19/82) Pregnancy rate 14% (29/201) 9.5% (19/200) Number of cows lost: AI but no pregnancy diagnosis 6 5 Number missing CL data 38 26 0 10 20 30 40 50 60 70 80 90 100 AICLCRPR Outcomes of InterestPercent Figure 2. The proportion of cows per group th at were inseminated (AI), returning to cyclicity, conception rate (CR), and pr egnancy rate (PR) for the Ovsynch group (white bars) and the CIDR group (black bars).
88 Cows in the Ovsynch group were enrolled in a timed insemination program and should ideally have had a 100% insemination rate, but due to different reasons beyond the control of this study, a proportion of these cows were not inseminated (82% inseminated; Table 4). Cows in the Ovsynch group that were not inseminated were not included in the denominator used for conception rate, but these cows were included in the denominator used for pregnancy rate. The pe rcentage of cows inseminated in the CIDR group is representative of the estrus detec tion rate for this herd. After cows were inseminated, a small number in each group were e ither sold from the he rd or not available for pregnancy determination. These numbers ar e presented as Cows lost in Table 4 and were subtracted from the total (denominator) used to calculate conception rate. Data for cows in the Ovsynch and CIDR groups were analyzed to determine the association between the explanatory variab les (treatment, DIM, BCS, parity, milk production, and time of year) and the outcome variables (likelihood to be inseminated, presence of a CL on Day 21, conception rate an d pregnancy rate). Lo gistic regression of the full model was used to determine expl anatory variables which were significantly associated with the outcome. Interactions between significant variables and treatment were tested for and included in the model, if found to be significant. The final model was chosen using backward elimination, forcing the effect of treatment to remain in the model and including all significant variables. Likelihood to be Inseminated In the CIDR group, 44% of cows were inseminated while 82 % of cows were inseminated in the Ovsynch group. The mean number of days between removal of the CIDR/PGF injection (Day 7) to observati on of estrus and insemination was 3.1 and ranged from 1 to 5 days.
89 The likelihood for a cow to be inseminated was significantly associated with the effect of treatment (Table 5; Figure 3). The odds for cows in the Ovsynch group to be inseminated were 5.6 times more than the odds for cows in the CIDR group (95% CI= 3.5-8.8; Table 5). Table 5. The risk of insemination for cows in the Ovsynch and CIDR groups adjusted for parity, DIM, BCS and milk produ ction on the day of diagnosis. Explanatory variables Odds Ratio 95% CI TreatmentOvsynch vs. CIDR 5.6 3.5 8.8 SeasonCool vs. Warm 1.4 0.8 2.3 ParityPP vs. MP 0.6 0.3 1.2 DIMEarly vs. Late 1.3 0.7 2.5 BCSLow vs. High 0.6 0.3 1.2 Milk production on the day of diagnosis Category 2 vs. 1 0.6 0.3 1.4 Category 3 vs. 1 0.8 0.3 2.1 Category 4 vs. 1 1.1 0.4 2.7 Ovsynch CIDR Figure 3. The proportion of cows inseminated (white) compared to cows not inseminated (black) for the Ovsynch and CIDR groups. Return to Cyclicity The presence of a CL on Day 21 (10 days after expected ovula tion) was used to determine the number of cows re turning to normal cyclicity. The percentage of cows in the Ovsynch and CIDR groups with a CL on Day 21 was 83% and 79%, respectively. The proportion of cows that were unavailabl e on Day 21 for determination of a CL is
90 presented in Table 4 (# missing CL data). Co ws unavailable for CL diagnosis were not included in the denominator used for calcu lating percent returning to cyclicity. The presence of a CL on Day 21 was significantly associated with the effect of treatment (Table 6). Adjusting for milk production, DI M, parity, season, and BCS, the odds for cows in the Ovsynch group to be diagnosed with the presence of a CL on Day 21 were 2.2 times more than the odds for cows in th e CIDR group (95% CI= 1.044.81; Table 6). No other explanatory variables were significantly associated with the presence of a CL on Day 21. Table 6. The risk for the presence of a CL on Day 21 for cows in the Ovsynch and CIDR groups adjusted for parity, DIM, BC S and milk production on the day of diagnosis. Explanatory variables Odds Ratio 95% CI TreatmentOvsynch vs. CIDR 2.2 1.04 4.81 SeasonCool vs. Warm 1.1 0.50 1.90 ParityPP vs. MP 0.5 0.26 1.30 DIMEarly vs. Late 0.6 0.27 1.34 BCSLow vs. High 0.6 0.27 1.34 Milk production on the day of diagnosis Category 2 vs. 1 1.2 0.41 3.56 Category 3 vs. 1 0.6 0.23 1.73 Category 4 vs. 1 1.3 0.41 4.20 The presence of a CL on Day 21 was signi ficantly associated with pregnancy (P= 0.05). The odds for cows with a CL on day 21 to become pregnant were 7.7 times more than for cows without a CL (95% CI= 0.97-58.8). Serum progesterone concentrations obtaine d on day 21 at the time of CL diagnosis in a subset of cows were used to confirm results obtained by palpation per rectum and ultrasonography, and to determine the accuracy of palpation per rectum in the diagnosis of a CL. In general, when cows were dia gnosed as not having a CL, serum progesterone concentrations were low (mean= 0.64ng/ml; Ta ble 7) with the exception of 3 outliers
91 (Figure 4). The majority of cows diagnosed with a CL had progesterone concentrations greater 3.0ng/ml, although values ranged fr om 0.027.7 (Table 7; Figure 4). Table 7. Descriptive statistics for progesterone values (ng/ml) based on the presence or absence of a CL CL No CL Median 3.09 0.02 Mean 2.87 0.64 Standard error 0.17 0.27 Standard deviation 2.00 1.38 Range 0.02-7.74 0.02-6.09 Figure 4. Box and whisker plots of progesterone values for cows with (CL) or cows without (No CL) a CL. A contingency table was used to evaluate the accuracy of palpa tion per rectum and ultrasound for the diagnosis of a CL (Table 8). In this table, the true presence of a CL was based on a progesterone value of > 1.0 ng/ml while the absence of a CL was based on a progesterone value < 1.0 ng/ml. Given this data and using established calculations (Slenning, 2001), the true prevalence of a CL was 63%, while the test prevalence (palpation per rectum) was 83%. The sensitivit y and specificity of palpation per rectum Progesterone ng/ml N o CL CL
92 and ultrasonography in the di agnosis of a CL were 97% and 39.7%, respectively. The positive and negative predictive values of palpation per rectum and ultrasonography in the diagnosis of a CL were 73.3% and 88.5 %, respectively. Table 8. Contingency table for the diagnosis of a CL based on palpation per rectum and U/S (RP+U/S) or progest erone value greater than or equal to 1.0ng/ml. Progesterone Value Presence of a CL based on RP+U/S > 1.0ng/ml < 1.0ng/ml Total CL 96 35 131 No CL 3 23 26 Total 99 58 Overall Total= 157 None of the cows with a clinical di agnosis of a CL and serum progesterone concentration < 1.0 ng/ml were pregnant at the following pregnancy diagnosis (Table 9). Of the remaining 95 cows with a CL and se rum progesterone concentration > 1.0 ng/ ml, 14 (15%) were pregnant (Table 9). Table 9. Progesterone level as it relates to pr egnancy in cows diagnosed with a CL on day 21. Serum Progesterone Not Pregnant Pregnant Total Low P4 (< 1.0ng/ml) 35 0 35 High P4 (>1.0ng/ml) 82 14 96 Total 117 14 131 Using logistic regression for only those cows with a CL, the association between serum progesterone concentration and pregna ncy was further defined. This analysis indicated that for each 1.0ng/ml increase in serum progesterone, the odds for a cow to be diagnosed pregnant increased by 1.6 times (95% CI= 1.01-2.47). The general linear model pr ocedure and least squares means were performed to determine which variables significantly infl uenced progesterone concentration on Day 21. Explanatory variables with a significant influence on progesterone concentration were: the presence of a CL on Day 21, BCS, milk production on the day of diagnosis, and
93 treatment (Table 10). As expected, cows di agnosed with the presence of a CL on Day 21 had significantly higher progest erone concentrations (P<0.0001; Table 10). Cows with a body condition score >2.75 had significantly hi gher progesterone concentrations (P= 0.047; Table 10). Cows with milk production > 28.8 kg had significantly higher progesterone concentrations (P= 0.017; Tabl e 10). Also, cows in the CIDR group had higher progesterone concentr ations than cows in the Ovsynch group (P=0.0452; Table 10), despite a decreased risk for the presen ce of a CL on Day 21 (P= 0.039; Table 6). The distribution of progestero ne concentrations in the Ovsynch and CIDR groups is presented in Figure 5. The least squares mean s for cows without a CL was similar to cows in the Ovsynch and CIDR groups at 0.81ng/ml and 0.56ng/ml, respectively, however, the least squares means were cons iderably different for cows these groups diagnosed with the presence of a CL at 2.63ng/ml and 3.32ng/ml, respectively. Table 10. Least squares means and standard error of the mean for progesterone concentrations and associations with the explanatory variables. Variable Progesterone LSM +/SEM(ng/ml) P value Treatment Ovsynch 1.64 +/0.24 0.045 CIDR 2.25 +/0.24 CL on Day 21 Presence 3.18 +/0.18 <0.0001 Absence 0.71 +/0.28 Milk Production on Day 0 < 28.8 kg 1.55 +/0.22 0.017 > 28.8 kg 2.34 +/0.26 BCS < 2.75 1.60 +/0.20 0.047 > 2.75 2.29 +/0.32 Parity 1st Lactation 2.19 +/0.26 0.125 > 2nd Lactation 1.69 +/0.22 DIM < 132 days 2.16 +/0.26 0.203 >132 days 1.73 +/0.23
94 Figure 5. Distribution of progesterone concentr ations (labeled P4 value; ng/ml) on Day 21 for cows in the Ovsynch (labeled Gr 1) and CIDR (labeled Gr 2) groups. Conception and Pregnancy Rates Conception and pregnancy rates for cows in the Ovsynch group were 18 and 14%, respectively, while conception and pregnancy rates for cows in the CIDR group were 23 and 9.5%, respectively (Table 4; Figure 2). Conception and pregnancy rates for cows in the Ovsynch group were not equal due to the difference caused by the 18% of cows that were lost between enrollment and pregnancy determination. All cows were included in the denominator for pregnancy rate while only those with an examination at 45 days were included in the denominator for conception rate.
95 With conception as the dependent variable, logistic regression of the full model and backward elimination showed no significant eff ect of treatment. The significant variables associated with conception rate were parity and milk production on the day of diagnosis. The odds for primiparous cows to conceive af ter treatment were 4.7 times more to than the odds for multiparous cows (95% CI= 1.5211.9; Table 11; Figure 6). The odds for cows with milk production in the 3rd quartile on the day of di agnosis (Category 3) to conceive were 12.78 times more than the odds for cows with milk production in Category 1 (95% CI= 1.42-114.9; Table 11; Figure 7). The odds for cows with milk production in Category 4 to conceive were 9.4 times more than the odds for cows with milk production in Category 1 (95% CI= 1.00488.4; Table 11; Figure 7). Table 11. The risk of conception for cows in the Ovsynch and CIDR groups adjusted for parity, DIM, BCS and milk produ ction on the day of diagnosis. Explanatory variables Odds Ratio 95% CI TreatmentOvsynch vs. CIDR 0.8 0.27 2.30 SeasonCool vs. Warm 1.4 0.72 2.92 ParityPP vs. MP 4.1 1.52 11.9 DIMEarly vs. Late 2.2 0.76 6.3 BCSLow vs. High 0.8 0.27 2.24 Milk production on the day of diagnosis Category 2 vs. 1 6.3 0.66 59.9 Category 3 vs. 1 12.8 1.42 114.9 Category 4 vs. 1 9.4 1.004 88.44
96 0 5 10 15 20 25 30 35 40 CRPRPercent 1st Lactation 2nd + Lactation Figure 6. Conception rate (CR) and pregna ncy rate (PR) divided by primiparous (1st Lactation; white) and multiparous (2nd + Lactation; black) for cows in the Ovsynch and CIDR groups. 0 5 10 15 20 25 30 35 40 45 Category 1Category 2Category 3Category 4 Milk Production% Pregnant Figure 7. Conception rate by milk production category for cows in the Ovsynch and CIDR group. The interactions of treatment x parity and treatment x milk production were not significant. Body condition score (BCS) and DIM were forced into the model, and milk production was removed to determine any associations between BCS, DIM and milk production. These variables still had no signi ficant effect on conception rate.
97 With pregnancy rate as the dependent vari able, logistic regression of the full model and backward elimination showed no significant e ffect of treatment. Similar to the results for conception rate, the variables significantly associated with pregnancy rate were parity and milk production. The odds for primiparous co ws to become pregnant after treatment were 3.7 times more than the odds for multip arous cows (95% CI= 1.25-10.6; Table 12; Figure 6). The odds for cows with milk production in the 3rd quartile on the day of diagnosis (Category 3) to become pregnant after treatment were 10.9 times more than the odds for cows with milk production in Category 1 (95% CI= 1.28-93.4; Table 12). The odds for cows with milk production in Category 4 to conceive after treatment were 9.3 times more than the odds for cows with milk production in Category 1 (95% CI= 1.0482.6; Table 12). Table 12. The risk factors associated with pr egnancy rate for cows in the Ovsynch and CIDR groups adjusted for parity, DI M, BCS and milk production on the day of diagnosis. Explanatory variables Odds Ratio 95% CI TreatmentOvsynch vs. CIDR 2.1 0.80 5.78 SeasonCool vs. Warm 1.6 0.84 3.10 ParityPP vs. MP 3.7 1.25 10.6 DIMEarly vs. Late 2.4 0.87 6.99 BCSLow vs. High 0.8 0.28 2.05 Milk production on the day of diagnosis Category 2 vs. 1 5.3 0.58 47.7 Category 3 vs. 1 10.9 1.28 93.4 Category 4 vs. 1 9.3 1.04 82.6
98 DISCUSSION The hypothesis of this study was that exoge nous progesterone with insemination at detected estrus would be a be tter therapeutic strategy for da iry cows with cystic ovarian disease than the Ovsynch protocol. The rati onale behind this hypothe sis was based on the current theory that cystic cows are proge sterone deficient, an d require progesterone exposure to correct the hypotha lamic defect predisposing to this condition (Gumen and Wiltbank, 2002). Based on this theory, it was expected that exposure to exogenous progesterone would up-regulate the number of estrogen receptors in hypothalamic neurons involved in transmission of the Gn RH surge, thus correcting the underlying hypothalamic defect. Further support for this hypothesis comes from recent work which has demonstrated that the Ovsynch protocol cau ses CL formation in only 50.9-70.9% of cows following the first GnRH injection (Hendricks 2004). This variable percentage occurs because the developmental stages of the unde rlying follicular waves in cystic cattle are unknown at the time of the 1st GnRH injection. The importance of the underlying follicle wave has been demonstrated in previous wo rk which indicates that the ideal time to initiate the Ovsynch protocol is during the earl y luteal phase of the estrous cycle, when a dominant follicle with ovulat ory capacity is present on the ovary (Vasconcelos et al., 1999; Moreira et al., 2000b). The 2nd GnRH injection of the Ovsynch protocol causes ovulation of the dominant follicle from the wave initiated after the first GnRH injection. Therefore, the success the 2nd GnRH injection is partially dependent on the success of 1st
99 GnRH injection in turning over the follicular wave. The Ovsynch protocol may not be the ideal treatment for cystic ovarian disease c onsidering its variability in the proportion of cows that ovulate following the 1st GnRH injection. This potentia l lack of ovulation limits the success of the 2nd injection and prolongs the period of progesterone deficiency for a proportion of these cows. However, the pr esence of a CL on Day 7 of the Ovsynch protocol increases the risk for pregnanc y in cystic cows (Hendricks, 2004). Given this information, it was speculated th at a protocol in which 100% of cows were exposed to progesterone, prior to a natural ovulation, would correct the underlying hypothalamic defect better than the Ovsynch protocol and improve the pregnancy rate. Contrary to our expectations, the results of the present study indicate that the CIDR protocol and the Ovsynch protoc ol were just as effective as therapeutic strategies for ovarian cysts with respect to conception and pregnancy rates, while the Ovsynch protocol was more effective than the CIDR protocol with respect to return to normal ovarian function. The likelihood for cows to be inseminate d in the CIDR and Ovsynch groups were 44% and 82%, respectively. The percentage of cows inseminated in the CIDR group (44%) is representative of the estrus detec tion rate for this farm. Ideally, the Ovsynch group should have had 100% insemination ra te but due to various reasons beyond the control of this study, 18% of cows in the Ovsynch group were not inseminated according to the protocol. This lack of compliance was considered part of the reality of performing a large field study in a privately owned herd and possibly reflects the true circumstances of many herds. Some reasons for the lack of insemination at the appropriate time may have included cows that were unavailable due to movement from their respective herd
100 (i.e. moved into the hospital ba rn) and cows not inseminated due to labor availability or management reasons. Considering this informa tion, it is also possible that a comparable proportion of cows in the CIDR group were ei ther not observed for estrus or may have been observed in estrus but not bred for sim ilar reasons to those listed above. If this is true it would be fair to speculate that the potential estrus detect ion rate for the CIDR group, under optimal conditions, could have been higher. Assuming there was some degree of lack of compliance in both groups, and to be fair to the CIDR group where this lack of compliance is hidden within the estrus detection rate, the to tal number of cows enrolled in each group was used as the denomin ator for pregnancy rate calculations. This was done even though it is common practice to include only cows inseminated when evaluating results for an Ovsynch protocol. Despite the proportion of cows not insemi nated in the Ovsynch group, the odds for cows to be inseminated in this group were 5.6 times more than the odds for cows in the CIDR group. Although this outcome was not unexpected when comparing a timedinsemination protocol to one relying on estrus detection, it emphasizes the major benefit of a timed-insemination protocol in a herd with less than optimal estrus detection. Furthermore, in a herd where the complian ce for the Ovsynch protocol may be higher, and with a similar estrus detection rate to the present herd, the Ov synch protocol could result in an even greater likelihood for co ws to be inseminated. Conversely, a higher estrus detection rate in the CIDR group would decrease the di fference in the likelihood to be inseminated between the two groups. Cows in the Ovsynch group were at an in creased risk for the presence of a CL following treatment compared to cows in the CIDR group (OR= 2.2; 95% CI=1.04
101 4.81). This indicates that more cows with cy stic ovarian disease will return to normal ovarian function following treatment with th e Ovsynch protocol than with the CIDR protocol. Although the percentage of cows in each group with a CL on Day 21 appeared similar (CIDR= 79%; Ovsynch =83%), the eff ect of treatment was significant when the model was adjusted for the influence of ot her variables (DIM, BCS, parity, season and milk production). The increased risk for cows in the Ovsynch group to have a CL on Day 21 may be partially explained by the action of the 2nd GnRH injection in causing a dominant follicle to ovulate. The CIDR protocol relies on a natural GnRH/LH surge and ovulation, which may or may not occur. When exogenous GnRH is administered, as in the Ovsynch protocol, an LH surge occurs in almost every case and when a dominant follicle is present, ovulation will occur in the majority of cases (Kaltenbach et al., 1974). In a study examining the ovarian response of cystic co ws to both a CIDR and Ovsynch protocol, one cow which did not ovulate following CI DR removal was successfully induced to ovulate by administration of Gn RH, thus emphasizing the va lue of a second GnRH in ensuring that ovulation occurs (Ambrose et al., 2004). Another possibility is the Gn RH given on Day 0 may have increased the risk for the presence of a CL on Day 21. Bartolome et al. (2005b) treated cystic cows with 4 protocols. All of those protocols incl uded a PGF injection on Day 7 and a GnRH injection on Day 9, 16h prior to timed inse mination, but with or without a GnRH injection on Day 0 and with or without the in sertion of a CIDR for 7 days beginning on Day 0. In that study, cystic cows which di d not receive a GnRH injection on Day 0 tended to be at a decreased risk for the pr esence of a CL on Day 17 compared to cows
102 which received GnRH (AOR=0.1; P=0.06). Although the present study used different protocols than those used by Bartolome et al (2005b), it is possible th at similar to those findings, the 1st GnRH given on Day 0 could have been the difference between the two protocols which led to more cows with a CL on Day 21. The percentage of cows returning to cy clicity in the Ovsynch group in this study (83%) was similar to that found by Ambrose et al (83% or 15/18; 2004) and slightly greater than that found by Fricke and Wiltba nk (73.1%; 1999). Return to cyclicity rates in cystic cows treated with various progesterone protocols have been reported to be between 61.5100% (Johnson and Ulberg, 1966; Zulu et al. 2003; Todorki et al., 2001; Calder et al., 1999). This variation is likely due to th e variable number of cows used, and the application of different protocols. The pr esent study included more cows and used a progesterone protocol slightly different than previous reports, yet achieved similar results (79%return to cyclicity). Return to cyclicity in the present study was determined by the presence of a CL on Day 21, or 11 days after expected ovulation. Serum progesterone valu es were obtained at the time of diagnosis to evaluate the accuracy of palpation per rectum and ultrasonography in determining the presence of a CL. Based on a serum progesterone value > 1ng/ml for the true presen ce of a CL, palpation per rectum and ultrasonography had a sensitivity, specificity, and positiv e predictive value (PPV) of 97%, 39.7%, and 73.3%, respectively. This indicate d that although palpation pe r rectum resulted in very few cows diagnosed with the absence of a CL while having high serum progesterone (false negatives); there were many more cows diagnosed with the presence of a CL while having low serum progesterone (false positives). The accuracy of palpation per rectum
103 for the diagnosis of a CL in the present st udy (PPV= 73.3%) is slightly less than previous reports (87%Archbald et al., 1992; 79%Ar chbald et al., 1993a). A possible explanation for this is that by day 11 or 12 of the estr ous cycle, when morphological evidence of a CL exists, a proportion of these may be unde rgoing functional regr ession. Although a normal CL usually begins regression around day 14 of the estrous cycl e (Stevenson, 1997), it may occur early in a proportion of cows w ith shorter estrous cycles. Furthermore, a significantly higher proportion of anovular cows (23%) have short luteal phases (<11 days) following the Ovsynch protocol than ovular cows (6%; Gume n et al., 2003). Thus, it is possible that because of their progesteron e deficient state, more cows in the Ovsynch group where experiencing short luteal phases an d regression by 11 days after ovulation. It was interesting that in the present study, 75% (6/8) of co ws with progesterone between 0.5 and 1.0ng/ml were diagnosed with a CL by palpation per rectum. This suggests there may have been a proportion of cows with mo rphological features of a CL, yet these corpora lutea were producing le ss progesterone than expected. Treatment significantly influenced progest erone concentrations obtained on Day 21 at the time of CL diagnosis. Despite the fact that cows in the Ovsynch group were more likely to be diagnosed with the presence of a CL (OR=2.2; 95% CI=1.04 4.81), cows in the CIDR group had a significantly higher m ean progesterone concentration (2.25 vs. 1.64ng/ml; P=0.045). There are two possible expl anations for this occurrence. Firstly, more cows in the CIDR group may have had intermediate levels of progesterone as a result of increased luteinization of the cyst, thus artificially raising progesterone concentrations in the absence of a CL. If this was the case, mean progesterone concentration for cows diagnosed with the presence of a CL woul d be similar in both
104 groups, while mean progesterone concentration for cows diagnosed with the absence of a CL would be higher in the CIDR group reflectin g luteinization of the cyst. Results of the least squares means for the interaction of tr eatment and CL indicat e the opposite effect. Cows in the Ovsynch and CIDR groups had si milar progesterone values when diagnosed with the absence of a CL (0.8 and 0.5ng/ml, respectively), yet when cows were diagnosed with the presence of a CL, the progesterone va lues were considerably more influenced by treatment (2.6 vs. 3.3ng/ml, respectively). Seco ndly, it is possible that corpora lutea induced by the Ovsynch protocol secrete less progesterone than corpora lutea resulting from a natural GnRH/LH surge following the withdrawal of exogenous progesterone. The finding of the present study that corpor a lutea induced by the Ovsynch protocol secrete less progesterone than corpora lut ea induced by the CIDR protocol can be explained by, but is not limited to, two possible physiological mechanisms. One possibility is that progesterone pretreatment in cystic cows may have a priming effect on the hypothalamus to improve the GnRH/LH surg e, tonic LH support for the CL, and/or alter the ability of the CL to respond to LH through changes in receptor dynamics. This theory is supported by work completed in the postpartum cow, which is also in a progesterone deficient state pr ior to ovulation and similar to cows with cystic ovarian disease. Postpartum cows pretreated with progesterone were less lik ely to experience a short luteal phase after GnRH induced ovul ation (Rutter et al., 1985), or a natural ovulation (Ramirez-Godinez et al., 1981), than cows without progest erone pretreatment. Similarly, postpartum beef cows which re ceived a progesterone implant prior to an induced ovulation had a profile of progesteron e concentration that was higher over the following 18 days than those not receiving a progesterone implant. Gumen et al. (2003)
105 found that anovular cows (cows without a CL) were more likely to have short luteal phases following treatment with the Ovsynch protocol than ovular cows. These results suggest the possibility that cows in the CIDR group, where a greater proportion were under the influence of progesterone prior to ovulation, may have had improved CL function and a greater serum progesterone concentration by day 11 after ovulation. The other possibility is that there was a difference betw een the ovulatory follicle of the Ovsynch and CIDR groups at the time of ovulation, perhaps in size or developmental stage, which altered the functionality of the future CL as it matured. The Ovsynch protocol results in ovulation of follicles which can be a variety of sizes and with a significant proportion < 11mm (26%) compared to the other sizes (Perry et al, 2005). In the same study, the majority of spontane ously ovulating follicles were > 13 mm. Furthermore, follicles induced to ovulate wh ich were less than 12.8 mm had significantly lower serum estradiol on the day of insemi nation, a slower rise in serum progesterone, and tended to have lower progesterone concen trations and increased embryonic mortality (Perry et al., 2005). In cows that ovulated spontaneously, follicle size had no effect on progesterone or fertility. Another study dete rmined that the post-ovulatory progesterone rise can be delayed by inducing luteolysis in the presence of a pre-ovulatory follicle <10 mm (Robinson et al., 2005). This resulted in a longer follicular phase prior to ovulation and a smaller ovulatory follicle than when lute olysis was induced in the presence of a follicle >10 mm. This smaller ovulatory follicle then formed a CL with decreased serum progesterone production, even though ovulati on occurred spontane ously. These results suggest that the physiological maturity of the preovulatory follicle (its steroidogenic capacity, number of granulosa cells, LH recep tors) can alter subsequent CL development
106 and progesterone production. It is possible that in the presen t study, the Ovsynch protocol induced the ovulation of less mature follicles compared to those which ovulated spontaneously in the CIDR protocol. This l ack of maturity in a proportion of follicles may have resulted in corpora lutea with decreased progesterone production, and a significant difference in the serum progester one concentrations by day 11 after ovulation between the Ovsynch and CIDR groups. Although the finding that cows in the CI DR group had higher progesterone on Day 21 is intriguing, it was not the intent of this study to evaluate the f actors influencing this outcome. Regardless, and perhaps more impor tantly, the influence of treatment on Day 21 progesterone concentration did not translate into an effect on fertility in the present study. This was despite the finding that for co ws diagnosed with the presence of a CL, each 1ng/ml increase in serum progesterone increased the odds for pregnancy by 1.6 times. It is possible that the difference in progesterone concentration would have also resulted in a significant effect of treatment on fertility if more cows had been used in this study or other factors influencing conception and pregnancy rates, such as compliance and heat detection efficienc y, were improved. Using the con ception rates obtained in this study, 805 cows per group would have been needed to demonstrate a significant difference. The general view th at increased CL progesterone production increases the risk for pregnancy is supported by other studies which have shown a positive association (Moore et al., 2005; Perry et al., 2005). The positive association between BCS and milk production with Day 21 progesterone concentration was not unexpected. It is possibl e that cows with a low body condition score or low milk pr oduction were in a state of ne gative energy balance at the
107 time of ovulation. Energy balance influences the amount of cholesterol available for steroidogenesis as well as the enzymatic activ ity. Thus, it follows that a negative energy balance could result in decreased progesterone production by the CL. There was no significant association between treatment and conception rate in the present study. The conception rate in the Ovsynch group (18%) was similar to that observed by Bartolome et al. (23.6%; 2000 and 16.7%; 2005b) and Hendricks (19.5%; 2004), but less than that observed by Fricke and Wiltbank (36.8%; 1999) using the same protocol in cows with cystic ovarian disease. The conception rates ob tained in response to the Ovsynch protocol in cows with cystic ov arian disease, in this and other studies, continues to be lower than that observed in cyclic cows (33%-Momcilovic et al., 1998; and 46-55%Pursley et al., 1995). The conception rate for cows in the CIDR group in the present study was 23%. This result is in agreement with the conception rates f ound by Zulu et al. (20%; 2003), using a PRID containing 1.55g of progesterone inserted for 12 days, and by Bartolome et al. (27.3%; 2005b), using a CIDR containi ng 1.38 g progesterone followed by timed insemination. Yet these conception rates are all considerably lower than that observed by Johnson and Ulberg (48.7-52.5%; 1966) using 14 daily injections of 50 or 100 mg of progesterone. This difference is likely due to the fact that cows in that study were observed for estrus over a pr olonged period (>45 days), theref ore allowing more time for recovery. Bartolome et al. (2005b) achieved a pregnancy rate of 37.5 % in cystic cows by inserting a CIDR for the first 7 days of the Ovsynch protocol. More studies are needed to determine if a CIDRsynch protocol improves c onception rate and is more cost effective than the standard Ovsynch protocol in cows with ovarian cysts.
108 Many studies have reported a 100% ovulation rate following progesterone treatment in cystic cows (Nanda et al., 1991; Todorki et al, 2001; Ca lder et al., 1999), yet there are few who have reported conception and pregnancy rates, and none of which have used a CIDR protocol similar to the present study. In studies where fertility results have been reported using various progesterone protocols (Johnson and Ul berg, 1966; Zulu et al., 2003; Bartolome et al., 2005b), there was not an obvious improvement compared to results reported using other treatment prot ocols, such as Ovsynch. The present study provided a direct comparison between a proge sterone based protocol and the Ovsynch protocol for the treatment of cystic ovarian disease in one large herd. The findings of the present study indicate that ther e was not a significant differen ce in the fertility of cystic cows treated with either the Ovsync h protocol or the CIDR protocol. The pregnancy rates in the Ovsynch a nd CIDR groups were 9.5% and 14 %, respectively. Contrary to the expectations of this study, the effect of treatment was not significantly associated with pregnancy rate. The pregnancy rate was defined as the number of cows pregnant divided by the total number enrolled in each group. As previously mentioned, the denominator in th e Ovsynch group included cows that were enrolled but not inseminated which is w hy; contrary to common approaches, the conception and pregnancy rates differ for this group. Success of treatment, with respect to pr egnancy rate, in the CIDR group depended on detection of estrus, which was 44% in this farm. If the insemination rate had been higher in the CIDR group, this may have tr anslated into a higher pregnancy rate. Therefore in a herd with a highe r rate of estrus detection, th e CIDR protocol may result in a higher pregnancy rate with potentially less labor and cow handling than the Ovsynch
109 protocol. However, in a herd with a lower ra te of estrus detection, the Ovsynch protocol is more likely to be a better treatment for ovarian cysts. Cows with in the top two categories of milk production on the day of diagnosis (Category 3 and 4) were more likely to conceive following insemination than cows with the lowest level of milk production (Table 11). Typically, higher milk production is associated with an increased incidence of cystic ovarian disease (Johnson et al., 1966; Lopez-Gatius et al., 2002; Bartlett et al., 1986) Yet, the association of milk production with response to treatment or spontaneous recovery has seldom been evaluated. The present results suggest that when cystic ova rian disease occurs in higher producing cows, they have a better response to treatment a nd improved fertility compared to cows with lower milk production. In partial disagreement with this finding, Lopez-Gatius et al. (2002) found that cows diagnosed with cy stic ovarian disease between days 43-49 postpartum with high milk pr oduction were less likely to spontaneously recover than cows with low milk production. Many studies have examined the relationshi p between milk yield and fertility in normally cyclic cows and, in contrast to the present study, found an antagonistic relationship. Despite this, there are still a handful of studies that disagree with the hypothesis that milk production and fertility are negatively correlated. These studies suggest there are other factors with a greater negative influence on fertility than milk production. Peters and Pursley (2002) found that cows with above average milk production had a greater pregnanc y rate in response to the Ovsynch protocol than cows with lower than average milk production (45.8 vs. 33.8%). Similar to the results of the present study, Buckley et al. ( 2003) found that cows with the 3rd highest estimated 305-d
110 cumulative solids corrected milk yield were 1.4 times more likely to become pregnant to the 1st service compared to cows with the lowest 305-d cumulative solids corrected milk yield. High milk production does not necessarily imply that the animal is also in a negative energy balance. Low producing cows of ten have lower dry matter intake and are at a greater risk for anestrous and low fertility than high producing cows (Staples and Thatcher, 1990). It is possible that cows with lower milk production are also experiencing a higher incidence of lameness, mastitis, and periparturie nt disorders. These other conditions may be affecting thei r fertility as well as their milk production. It has been speculated that improved reproduction in hi gh-producing cows proba bly reflects better feeding, healthier cows and improved reproductive management (Lucy, 2001). A direct comparison between this and othe r studies with respect to milk production should be done with caution as many studies have analyzed milk production over a longer period, such as the month of diagnosis or 305day milk yield. The values obtained in the present study reflect the milk production of one day only, the day of diagnosis, and do not indicate how milk production wa s changing at the time of dia gnosis or the potential 305-d yield for the cow. Despite this the variable used for milk production in the present study had a significant influence on progesterone concentration, conception and pregnancy rates. In the present study, it was a very important explanatory variable for the response to treatment and subsequent fertility of cows with cystic ovarian disease. An interesting finding in the present study was that primiparous cows were more likely to be diagnosed pregnant after treatm ent than multiparous cows. Primiparous cows are generally considered to be at a lower ri sk for cystic ovarian disease than multiparous cows (Whitmore et al., 1974; Bartlett et al., 198 6; Hoojer et al., 2001). It is important to
111 remember that relative risk for the condition does not necessarily relate to how well they will respond to treatment. In partial agreemen t with the present study, Lopez-Gatius et al. (2002) found that primiparous cows diagnosed with cystic ovarian disease between days 43-49 postpartum were more likely to spontaneously recover than multiparous cows, such that a 1 unit drop in lactation number was asso ciated with a 1.4 increased probability of recovery. Similarly, of cows without ovarian cysts that were treated with the Ovsynch protocol, primiparous cows were more li kely to conceive than cows with > 3 lactations (Peters and Pursley, 2002). In contrast to these findings and to thos e of the present study, Bartolome et al. (2003) found that in cystic cows pretreated with or without bST and GnRH prior to the Ovsynch protocol, primiparous cows were 0.4 times less likely to conceive than multiparous cows. It is possible that diffe rences in response to treatment between primiparous and multiparous cows in this and other studies may be partially explained by management practices involving primiparous cows. Age at calving, nutrition, stress and housing environment differences between he rds may partially explain the variable response to treatment and fertility among prim iparous cows in different studies. Again, comparisons and contrasts between this and ot her studies should be done with caution as many other studies divided cows into 1st, 2nd, and 3rd+ lactations while the present study used only 1st and 2nd+ lactations. In studies th at have divided cows by 1st, 2nd, and 3rd+ lactations, many have observed that 1st lactation cows have different results than cows in the 3rd+ lactation, but were not significantly different from 2nd lactation cows. In the present study, season and BSC were not significantly associated with the response to treatment. As with milk producti on and parity, there is evidence for an
112 association between season, body condition scor e and the incidence of cystic ovarian disease. In one study, cows calving in the su mmer months were 2.6 times more likely to develop cysts than those calvi ng in the winter (Lopez-Gat uis et al., 2002). However, other authors have found no significant effect of season on the incidence of cysts (Bartlett et al., 1986). The effect of season on respons e to treatment and fe rtility is less often reported in cystic cattle than in normal cattl e. Using the periods of October to December and January to May, Bartolome et al. (2000) did not find an effect of season on the fertility of cystic cows following different treatment protocols in a Florida dairy herd. Typically in subtropical environments such as Florida, fertility is significantly decreased during the summer months (AlKatanani et al., 1999; Donova n et al., 2003; De Renesis and Scaramuzzi, 2003). The lack of seasonal effect on response to treatment in the present study may be due to successful heat stress management with in the study herd. It also may be that cystic cattle are at a decrea sed risk for the negative effects of heat stress on fertility A low body condition score, or a change in body condition score over time, has been associated with an increased risk fo r cystic ovarian disease as well as other metabolic diseases. A 1 unit change in BCS in cows between day 60 prepartum and parturition resulted in an 8.4 times increased risk for deve loping cystic ovarian disease between days 43 and 49 postpartum (Lopez-Ga tius et al., 2002). In another study, there was a significant negative linear relationshi p between cows with low BCS (<3.25) and the percentage of anovular cows, such that th e percentage of anovular cows decreased as the BCS increased (Gumen et al., 2003). Furt hermore, in the same study anovular cows with smaller follicles (<14mm) also tended to have the lowest BCS (< 2.5). The
113 association of BCS with fertility in cows w ithout ovarian cysts has been evaluated more frequently than its association with fertility in cows with ovarian cysts. In the present study, the only significant relationship i nvolving BCS was a positive association with serum progesterone on Day 21 (P=0.047; Table 10). In a study combining the results of cows w ith ovarian cysts and cow in proestrus, cows with a BCS >3.0 had an increased risk for the presence of a CL 7 days after timed insemination compared to cows with a BCS < 3.0 (Bartolome et al ., 2005b). In the same study, cows with a BCS > 3.0 also tended to have an increased risk for pregnancy on days 30 and 55 after insemination. In another study us ing normally cyclic cows, cows with low BCS ( < 2.5) were less likely to be inseminated dur ing the first 3 weeks after the voluntary waiting period than cows with higher BCS, and cows that reached a low nadir BCS (< 2.5) had a reduced likelihood for pregnancy to th e first service (Buckl ey et al., 2003). It is possible that the lack of association between BCS and fer tility in the present study was because BCS was only recorded on one occasio n, on the day of diagnosis, which does not indicate the direction of change or the energy balance of th e cow at that time. If detected, a declining BCS would indicate a negative energy balance an d would be more likely to negatively affect fertility. After enrollment, more cows were lost than expected throug hout the study period. These losses were the result of death, diseas e and culling decisions. A total of 11 cows were inseminated but lost prior to pregna ncy diagnosis and a total of 64 cows were unavailable on Day 21 for the diagnosis of a CL (Table 4). Many of the cows unavailable for CL diagnosis were in the hospital barn or missed being separated on the day of the reproductive examination. It was anticipated th at a small percentage of cows would be
114 lost after enrollment. It was for this reason that more cows were originally enrolled (n=401) than required to show a significant difference base d on sample size calculations (n=260). Due to the additional cows enroll ed, it is unlikely that the cow losses experienced were detrimental to the results. Compliance within the Ovsynch protocol ( 18% were not inseminated) was less than expected. Some potential reasons for the lack of insemination have been previously mentioned. If compliance had been better with in the Ovsynch protocol, the results would favor the Ovsynch protocol. For example, if compliance was 100% and conception rate was unchanged, pregnancy rates for the Ovs ynch and CIDR groups would have been 18 and 9.5%, respectively. There would be a st atistically significan t difference between these two pregnancy rates and cows in the Ovsynch group would have had an increased likelihood to be diagnosed pregnant once enrolled in the study. In conclusion, fertility was not statistical ly different between cystic cows treated with either the Ovsynch protocol or e xogenous progesterone and inseminated at an induced estrus. Cows treated with the Ovsync h protocol were more likely to return to normal cyclicity than cows treated with the CIDR protocol. Despite a decreased likelihood for cows in the CIDR group to be di agnosed with the presence of a CL, cows in this group had higher average progesterone concentrations 11 da ys after ovulation, and corpora lutea induced by the CIDR protocol secreted more progesterone than those induced by the Ovsynch protocol. Primipar ous cows had an increased likelihood for conception than multiparous cows, regardless of treatment. Cows in the 3rd and 4th quartiles of milk production on the day of dia gnosis were more likely to conceive than those in the 1st quartile, regardless of treatment.
115 SUMMARY AND CONSLUSIONS A total of 401 lactating dairy cows with ova rian cysts were treated with either the Ovsynch protocol or a protocol using a CI DR. Cows in the Ovsynch group were timedinseminated while cows in the CIDR groups we re inseminated at an induced estrus. The proportion of cows in each group that were in seminated, returned to normal cyclicity, and became pregnant was compared using l ogistic regression. Serum progesterone concentration was determined 11 days afte r expected ovulation (at the time of CL diagnosis). Cows in the Ovsynch group were more likely to be inseminated and more likely to return to normal cyclicity compared to cows in the CIDR group. However, cows in the CIDR group had higher progesterone concentrations 11 days after expected ovulation compared to cows in the Ovsync h group. There was no difference between the two treatment protocols in eith er the conception or pregnanc y rates. Factors which were significantly associated with fertility were parity and milk production on the day of diagnosis. Primiparous cows and cows with higher milk production were more likely to become pregnant than multiparous cows or cows with lower milk production on the day of diagnosis. In conclusion, there was no difference in fe rtility between cows with ovarian cysts treated with either the Ovsynch or CIDR pr otocols. However, cows treated with the Ovsynch protocol were mo re likely to return to normal cyclicity.
116 LIST OF REFERENCES Adams TE, Kinder JE, Chakraborty PK, Esterg reen VL, Reeves JJ. Ewe luteal function influenced by pulsatile administration of synthetic LHRH/FSHRH. Endocrinology 1975;97:1460-1467. Adams GP, Matteri RL, Kastelic JP, Ko JCH, Ginther OJ. Association between surges of follicle-stimulating hormone and the emergence of follicular waves in heifers. J Repod Fert 1992;94:177-188. Agresti A. An introduction to categorical da ta analysis. New York: John Wiley & Sons, 1996. p.103-132. Al-Katanani YM, Webb DW. Hansen PJ. Factor s affecting seasonal variation in 90-day nonreturn rate to first service in lactating holstein cows in a hot climate. J Dairy Sci 1999;82:2611-2616. Ambrose DJ, Schmitt EJP, Lopes FL, Mattos RC, Thatcher WW. Ovarian and endocrine responses associated with the treatment of cystic ovarian follicles in dairy cows with gonadotropin releasing hor mone, and prostaglandin F2 with or without exogenous progesterone. Can Vet J 2004;45:931-937. Anderson L. Intracellular mechanisms tr iggering gonadotropin secretion. Rev Reprod 1996;1:193-202. Anderson LH, Day ML. Acute progesterone ad ministration regresses persistent dominant follicles and improves fertility of cattle in which estrus was synchronized with melengestrol acetate. J Anim Sci 1994;72:2955-2961. Ando T, Kamimura S, Hamana K, Watanabe G, Taya K. Treatment at CIDR insertion influences ovarian follicularr dynamics in japanese Black cows. J Vet Med Sci 2005;63:275-280. Archbald LF, Norman SN, Tran T, Lyle S, Thomas PGA. Does GnRH work as well as GnRH and PGF2 in the treatment of ovarian follicular cysts. Vet Med 1991;86:1037-1040. Archbald LF, Thatcher WW. Ovarian follicul ar dynamics and management of ovarian cysts. In: VanHorn HH, Wilcox CJ ed itors. Large Dairy Herd Management. Champaign IL: ADSA, 1992. p.199-208.
117 Archbald LF, Tran T, Massey R, Klapstein E. Conception rates in dairy cows after timedinsemination and simultaneous treatment with gonadotropin releasing hormone and/or prostaglandin F2 al pha. Theriogenology 1992;37:723-731. Archbald LF, Risco C, Constant S, Tran T, Klapstein E, Elliot J. Estrus and pregnancy rate of dairy cows given one or two dose s of prostaglandin F2 alpha 8 or 24 hours apart. Theriogenology 1993a;40:873-884. Archbald LF, Sumrall DP, Tran T, Klapstei n E, Risco C, Chavatte P. Comparison of pregnancy rates of repeat-breeder da iry cows given gonadotropin-releasing hormone at or prior to the tie of insemination. Theriogenology 1993b;38:10811091. Armstrong DG, Baxter G, Gutierrez CG, H ogg CO, Glazyrin AL, Campbell BK, Bramley TA, Webb R. Insulin-like growth factor binding protein -2 and -4 messenger ribonucleic acid expression in bovine ovarian follicles: effect of gonadotropins and developmental status. Endocrinology 1998;139:2146-2154. At-Taras EE, Spahr SL. Detection and characte rization of estrus in dairy cattle with an electronic heatmount detector and an electronic activity tag. J Dairy Sci 2001;84:792-798. Baird DT. Pulsatile secretion of LH and ovari an estradiol during the follicular phase of the sheep estrous cycle. Biol Reprod 1978;18:359-364. Bartlett PC, Ngategize PK, Kaneene JB, Ki rk JH, Anderson SM, Mather EC. Cystic follicular disease in Michigan Holstein-F riesian cattle: Incidence, descriptive epidemiology and economic impact. Prev Vet Med 1986;4:15-33. Bartolome JA, Archbald LF, Morresey P, Hern andez J, Tran T, Kelbert D, Long K, Risco CA, Thatcher WW. Comparison of synchronization of ovulation and induction of estrus as therapeutic strategi es for bovine ovarian cysts in the dairy cow. Theriogenology 2000;53:815-825. Bartolome J, Hernandez J, Landaeta A, Kellema n A, Sheerin P, Risco CA, Archbald LF. The effect of interval from day of admi nistration of bovine somatotropin (bST) to synchronization of ovulation and timed-in semination on conception rate of dairy cows with and without ovarian cy sts. Theriogenology 2002;57:1293-1301. Bartolome J, Hernandez J, Sheerin P, Luznar S, Kelbert D, Thatcher WW, Archbald LF. Effect of pretreatment with bovine somatotropin (bST) and/or gonadotropinreleasing hormone (GnRH) on conception ra te of dairy cows with ovarian cysts subjected to synchronization of ovulati on and timed insemination. Theriogenology 2003;59:1991-1997.
118 Bartolome JA, Santos JEP, Pancarci SM, Melendez P, Arteche ACM, Hernandez O, Archbald LF, Trigg T, Thatcher WW. I nduction of ovulation in nonlactating dairy cows and heifers using different doses of a deslorelin implant. Theriogenology 2004;61:407-419. Bartolome JA, Sozzi A, McHale J, Melendez P, Arteche ACM, Silvestre FT, Kelbert D, Swift K, Archbald LF, Thatcher WW. Resynchronization of ovulation and timed insemination in lactating dairy cows, II: assigning protocols according to stages of the estrous cycle, or presence of ovarian cysts or anestrus. Theriogenology 2005a;63:1628-1642. Bartolome JA, Sozzi A, McHale J, Swift K, Kelbert D, Archbald LF, Thatcher WW. Resynchronization of ovulation and timed in semination in lactating dairy cows III. Administration of GnRH 23 days post AI and ultrasonography for nonpregnancy diagnosis on day 30. Th eriogenology 2005b;63:1643-1658. Battaglia DF, Beaver AB, Harris TG, Tanhe hco E, Viguie C, Karsch FJ. Endotoxin disrupts the estradiol-induced luteiniz ing hormone surge: Interference with estradiol signal reading, not surg e release. Endocrinology 1999;140:2471-2479. Battaglia DF, Krasa HB, Padmanabhan V, Viguie C, Karsch FJ. Endocrine alterations that underlie endotoxin-induced disruption of the follicular phase in ewes. Biol Reprod 2000;62:45-53. Beg MA, Bergfelt DR, Kot K, Ginther OJ. Fo llicle selection in cattle: dynamics of follicular fluid factors during development of follicle dominance. Biol Reprod 2002;66:120-126. Beg MA, Meira C, Bergfelt DR, Ginther OJ. Role of oestradiol in growth of follicles and follicle deviation in heifers. Reproduction 2003;125:847-854. Belchetz PE, Plant TM, Nakai Y, Keogh EJ, Knobil E. Hypophysial responses to continuous and intermittent delivery of hypothalamic gonadotropin-releasing hormone. Science 1978;602:631-632. Bergfeld EG, Kojima FN, Cupp AS, Wehrman ME Peters KE, Mariscal V, Sanchez T, Kinder JE. Changing dose of progesterone re sults in sudden changes in frequency of luteinizing hormone pulses and secreti on of 17 beta-estradiol in bovine females. Biol Reprod. 1996;54:546-553. Bierschwal CJ. A clinical study of cystic conditions of the bovine ovary. J Am Vet Med Assoc 1966;149:1591-1595. Bierschwal CJ, Gaverick HA, Martin CE, Youngquist RS, Cantley TC, Brown MD. Clinical response of dairy cows with ovarian cysts to GnRH. J Anim Sci 1975;41:1660-1665.
119 Blache D, Batallier M, Fabre-Nys C. Oestr ogen receptors in the preoptico-hypothalamic continuum: Immunohistochemical study of th e distribution and cell density during induced oestrous cycle in ovariectomi zed ewe. J. Neuroendocrinology 1994;6:329339. Bo GA, Pierson RA, Mapletoft RJ. The effect of estradiol valerate on follicular dynamics and superovulatory response in cows with Synchro-mate-B implants. Theriogenology 1991;36:169-183. Bosberry S, Dobson H. Periparturient di seases and their effect on reproductive performance in five dairy herds. Vet Rec 1989;124:217-219. Bosu WTK, Peter AT. Evidence for a role of in trauterine infections in the pathogenesis of cystic ovaries in postpartum dair y cows. Theriogenology 1987;28:725-736. Britt JH, Harrison DS, Morrow DA. Frequency of ovarian follicular cysts, reasons for culling, and fertility in Holstein-Fri esian cows given gonadotropin-releasing hormone at two weeks after part urition. Am J Vet Res 1977;38:749-751. Britt JS, Gaska J. Comparison of two estrus synchronization programs in a large, confinement-housed dairy herd. J Am Vet Med Assoc 1998;212:210-212. Buckley F, OSullivan K, Mee JF, Evans RD Dillon P. Relationship among milk yield, body condition score and reproduction in spri ng-calved Holstein-Friesians. J Dairy Sci 2003;86:2308-2319. Burcelin R, Thorens B, Glauser M, Gailla rd RC, Pralong FP. Gonadotropin-releasing hormone secretion from hypothalamic neurons: stimulation by insulin and potentiation by leptin. Endocrinology 2003;144:4484-4491. Burke JM, De la Sota RL, Risco CA, Staples CR, Schmitt EJ-P, Thatcher WW. Evaluation of timed insemination using a gonadotropin-releasing hormone agonist in lactating dairy cows J Dairy Sci 1996;79:1385-1393. Busato A, Romagnoli S, Kupfer U, Rossi GL, Bestetti GE. LH, FSH, PRL and ACTH cells in pituitary glands of cows w ith ovarian cysts. Theriogenology 1995;44:233246. Calder MD, Salfen BE, Bao B, Youngquist RS, Garverick HA. Administration of progesterone to cows with ovarian follicula r cysts results in a reduction in mean LH and LH pulse frequency and initiates ovul atory follicular growth. J Anim Sci. 1999;77:3037-3042. Calder MD, Manikkam M, Salfen BE, Youngquist RS, Lubahn DB, Lamberson WR, Gaverick HA. Dominant bovine ovarian follic ular cysts express increased levels of messenger RNA for luteinizing hormone receptor and 3 -hydroxysteroid dehydrogenase 4, 5 isomerase compared to normal dominant follicles. Biol Reprod. 2001;65:471-476.
120 Campbell RE, Grove KL, Smith MS. Gonadotr opin-releasing hormone neurons coexpress orexin 1 receptor immunoreac tivity and receive direct c ontacts by orexin fibers. Endocrinology 2003;144:1542-1548. Cantley TC, Gaverick HA, Bierschwal CJ, Martin CE, Youngquist RS. Hormonal responses of dairy cows with ovarian cysts to GnRH. J Anim Sci 1975;41:16661673. Caraty A, Evans NP, Fabre-Nys CJ, Karsch FJ. The preovulatory gonadotropin-releasing hormone surge: a neuroendocrine signal for ovulation. J Reprod Fert 1995;Suppl 49:245-255. Caraty A, Fabre-Nys C, Delaleu B, Locatel li A, Bryuneau G, Karsch FJ, Herbison A. Evidence that the mediobasal hypothalamus is the primary site of action of estradiol in inducing the preovulatory gona dotropin releasing hormone surge in the ewe. Endocrinology 1998;139:1752-1760. Caraty A, Skinner DC. Progesterone priming is essential for the full expression of the positive feedback effect of estradiol in inducing the preovulatory gonadotropinreleasing hormone surge in th e ewe. Endocrinology 1999;140:165-170. Carroll DJ, Pierson RA, Hauser ER, Grumme r RR, Combs DK. Vari ability of ovarian structures and plasma progesterone profile s in dairy cows with ovarian cysts. Theriogenology 1990;34:349-370. Cavalieri J, Hepworth G, Fitzpatrick LA. Comparison of two estr us synchronization and resynchronization treatments in lact ating dairy cows. Theriogenology 2004;62:729747. Chabbert-Buffet N, Skinner DC, Caraty A, Bouchard P. Neuroendocrine effects of progesterone. Steroids 2000;65:613-620. Chappell PE, Lydon JP, Conneely OM, OMalley BW, Levine JE. Endocrine defects in mice carrying a null mutation for the progesterone receptor gene. Endocrinology 1997;138:4147-4152. Chappell PE, Levine JE. Stimulation of gondaotropin-releasing hormone surges by estrogen. I. Role of hypothalamic pr ogesterone receptors. Endocrinology 2000;141:1477-1485. Chappell PE, Schneider JS, Kim P, Xu M, Lydon JP, OMalley BW, Levine JE. Absence of gonadotropin surges and gonadotropinreleasing hormone self priming in ovariectomized, estrogen-treated, progesterone receptor knockout mice. Endocrinology 1999;140:3653-3658. Christian RE, Casida LE. The effects of proge sterone in altering the estrus cycle of the cow. J Anim Sci 1948;7:540.
121 Clarke IJ, Cummins JT, Crowder ME, Nett TM. Pituitary receptors for gonadotropinreleasing hormone in ovariectomized-hypotha lamo pituitary disconnected ewes: I. Effect of changing frequency of gonadot ropin-releasing hormone pulses. Biol Reprod 1987;37:749-754. Cook DL, Parfet JR, Smith CA, Moss GE, Younguist RS, Gaverick HA. Secretory patterns of LH and FSH during deve lopment and hypothalamic and hypophysial characteristics following development of st eroid-induced ovarian follicular cysts in dairy cows. J Reprod Fert 1991;91:19-28. Cook DL, Smith CA, Parfet JR, Younguist RS, Brown EM, Gaverick HA. Fate and turnover rate of ovarian follicular cysts in dairy cattle. J Reprod Fert 1990;90:3746. Cooper MJ. Control of oestrous cycles of heifers with a s ynthetic prostaglandin analogue. Vet Rec 1974;95:200-203. Couse JF, Korach KS. Estrogen receptor null mice: What have we learned and where will they lead us? Endocrine Rev. 1999;20:358-417. Couse JF, Yates MM, Sanford R, Nyska A, N ilson JH, Korach KS. Formation of cystic ovarian follicles associated with elevat ed luteinizing hormone requires estrogen receptor. Endocrinology 2004;145:4693-4702. Davis TL, Mussard ML, Jimenez-Severiano H, Enright WJ, Kinder JE. Chronic treatment with an agonist of gonadotr opin-releasing hormone enha nces luteal function in cattle. Biol Re prod 2003;69:398-403. De Rensis F, Scaramuzzi RJ. Heat stress a nd seasonal effects on reproduction in the dairy cowsa review. Theriogenology 2003;60:1139-1151. De Silvia M, Reeves JJ. Hypothalamic-pitu itary function in chronically cystic and regularly cycling dairy co ws. Biol. Reprod. 1988;38:264-269. de Vries A, Conlin BJ. Economic value of timely determination of unexpected decreases in detection of estrus using cont rol charts. J Dairy Sci 2003;86:3516-3526. de Vries A. Economics of delayed replacement when cow performance is seasonal. J Dairy Sci 2004;87:2947-2958. Dobson H. Plasma gonadotropins and oestradiol during estrus in the cow. J Reprod Fertil 1977;52:51-53. Dobson H, Alam MG. Preliminary investigations into the endoc rine systems of subfertile cattle: location of a comm on lesion (rate-limiting step ). J Endocrinol 1987;113:167171.
122 Dobson H, Campbell BK, Gordon B, Scaramuz zi RJ. Endocrine activity of induced persistent follicles in sh eep. Biol. Reprod. 1997;56:208-213. Donovan GA, Bennett FL, Springer FS. Factors a ssociated with first service conception in artificially inseminated nulliparou s Holstein heifers. Theriogenology 2003;60:67-75. Douthwaite R, Dobson H. Compar ison of different methods of diagnosis of cystic ovarian disease in cattle and an assessment of its treatment with a progesterone releasing intravaginal device. Vet Rec 2000;147:355-359. doValle ER, Cruz LC, Kesler DJ. Gonadotro pin-releasing hormone enhances the calving rate of beef females administered no rgestomet and alfaprosol for estrus synchronization. J Amin Sci 1997;75:897-903. El-Zarkouny SZ, Cartmill JA, Hensley BA, Stev enson JS. Pregnancy in dairy cows after synchronized ovulation regimens with or without presynchronization and progesterone. J Dairy Sci 2004;87:1024-1037. Erb HN, White ME. Incidence rate s of cystic follicles in Holstein cows according to 15day and 30-day intervals. Cornell Vet 1981;71:326-331. Evans NP, Dahl GE, Caraty A, Padmanabhan V, Thrun LA, Karsch FJ. How much of the gonadotropin-releasing hormone (GnRH) surg e is required for generation of the luteinizing hormone surge in the ew e? Duration of the endogenous signal. Endocrinology 1996;137:4730-4737. Evans ACO, Komar CM, Wandji S-A, Fortune JE. Changes in androgen secretion and luteinizing hormone pulse amplitude are associated with the recruitment and growth of ovarian follicles during the luteal phase of th e bovine estrous cycle. Biol Reprod 1997;57:394-401. Farin PW, Estill CT. Infertility due to abnormalit ies of the ovaries in cattle. Vet Clinics of North America: Food Animal Practice 1993;9:291-308. Farin PW, Younquist RS, Parfet JR, Gaveric k, HA. Diagnosis of luteal and follicular ovarian cysts in dairy cows by sector scan ultrasonography. Theriogenology 1990;34:636-642. Farin PW, Younquist RS, Parfet JR, Gaveric k, HA. Diagnosis of luteal and follicular ovarian cysts by palpation per rectum and linear arra y ultrasonography in dairy cows. J Am Vet Med Assoc 1992;200:1085-1089. Fernandez-Lima O, Murphy BD, Manns JG. S ynchronization of preovulatory LH surges after PGF2 in beef cattle: The effect of Dala-GnRH and testosterone benzoate. Theriogenology 1977;7:73-79.
123 Ferguson JD, Galligan DT and Thomsen N. Pr incipal descriptors of body condition score in Holstein cows. J Dairy Sci 1994;77:2695-2703. Fleischer P, Metzner M, Hoedemaker M, Slos arkova S, Skrivanek M. Clinical disorders in Holstein cows: Incidence and associati ons among lactational risk factors. Acta Vet Brno 2001;70:157-165. Folman Y, Kaim M, Herz Z, Rosenbe rg M. Comparison of methods for the synchronization of estrous cycles in dair y cows. 2. Effects of progesterone and parity on conception. J Dairy Sci 1990;73:2817-2825. Fortune JE. Ovarian follicular growth and development in mammals. Biol Reprod 1994;50:225 232. Fricke PM, Wiltbank MC. Effect of milk production on the incidence of double ovulation in dairy cows. Theriogenology 1999;52:1133-1143. Garverick HA. Ovarian follicular dynamics and endocrine profiles in cows with ovarian follicular cysts. In: Howard JL, Smith RA (Eds.), Current Veterinary Therapy, Food Animal Practice. Philadelphia: WB Saunders Company, 1999; pp.577-580. Gazal OS, Leshin LS, Stanko RL, Thomas MG, Keisler DH, Anderson LL, Williams GL. Gonadotropin-releasing hormone secretion into third vent ricle cerebrospinal fluid of cattle: correspondence with the tonic a nd surge release of luteinizing hormone and its tonic inhibition by suckling and neuropeptid e Y. Biol Reprod 1998;59:676683. Gibbons JR, Wiltbank MC, Ginther OJ. Relatio nship between follicular development and the decline in the follicle-stimulating hor mone surge in heifers. Biol Reprod 1999;60:72-77. Ginther OJ, Wiltbank MC, Fricke PM, Gibbons JR, and Kot K. Selection of the dominant follicle in cattle. Biol Reprod 1996;55:1187-1194. Ginther OJ, Bergfelt DR, Kulick LJ, Kot K. Sel ection of the dominant follicle in cattle: establishment of follicle deviation in le ss than 8 hours through depression of FSH concentrations. Theriogenology 1999;52:1079-1093. Ginther OJ, Bergfelt DR, Kulick LJ, Kot K. Sel ection of the dominant follicle in cattle: role of estradiol. Biol Reprod 2000a;63:383-389. Ginther OJ, Bergfelt DR, Kulick LJ, Kot K. Sel ection of the dominant follicle in cattle: role of the two-way functional coupling between follicle-stimulating hormone and the follicles. Biol Reprod 2000b;62:920-927. Ginther OJ, Bergfelt DR, Beg MA, Kot K. Effe ct of LH on circulat ing oestradiol and follicular fluid factor concentrations during follicle deviation in cattle. Reproduction 2001a;122:103-110.
124 Ginther OJ, Bergfelt DR, Beg MA, Kot K. Folli cle selection in cattle: role of luteinizing hormone. Biol Reprod 2001b;64:197-205. Ginther OJ, Bergfelt DR, Beg MA, Meira C, Kot K. In vivo effects of an intrafollicular injection of insulin-like growth factor 1 on the mechanism of follicle deviation in heifers and mares. Bi ol Reprod 2004;70:99-105. Glister C, Tanetta DS, Groome NP, Knight PG Interactions between follicle stimulating hormone and growth factors in modulating se cretion of steroids and inhibin-related peptides by nonluteinized bovine granul ose cells. Biol Reprod 2001;65:1020-1028. Gong JG, Campbell BK, Bramley TA, Gutierrez CG, Peters AR, Webb R. Suppression in the secretion of follicle-stimulating hormone and luteinizing hormone, and ovarian follicle development in heifers continuously infused with a gonadotropin-releasing hormone agonist. Bi ol Reprod 1996;55:68-74. Goodman RL, Karsch FJ. Pulsatile secreti on of luteinizing hormone: differential suppression by ovarian steroids Endocrinology 1980;107:1286-1290. Goodman RL, Coolen LM, Anderson GM, Hardy SL, Valent M, Connors JM, Fitzgerald ME, Lehman MN. Evidence that dynorphi n plays a major role in mediating progesterone negative feedback on gona dotropin-releasing hormone neurons in sheep. Endocrinology 2004;145:2959-2967. Grimus RL, Wise ME. Progesterone directly inhibits pituitary luteinizing hormone secretion in an estradiol-dependent manner. Biol Reprod 1992;46:710-714. Grhn YT, Eicker SW, Ducrocq V, Hertl JA. Effect of diseases on the culling of Holstein dairy cows in New York St ate. J Dairy Sci 1998;81:966-978. Gumen A, Sartori R, Costa FMJ, Wiltbank MC A GnRH/ LH surge without subsequent progesterone exposure can induce developm ent of follicular cysts. J Dairy Sci 2002;85:43-50. Gumen A, Wiltbank MC. An alteration in the h ypothalamic action of estradiol due to lack of progesterone exposure can cause follic ular cysts in cattle. Biol Reprod 2002;66:1689-1695. Gumen A, Guenther JN, Wiltbank MC. Follic ular size and response to ovsynch versus detection of estrus in anovular and ov ular lactating dairy cows. J Dairy Sci 2003;86:3184-3194. Gumen A, Wiltbank MC. Length of progesterone exposure needed to resolve large follicle anovular condition in dairy cows. Theriogenology. 2005;63:202-218. Halter TB, Hayes SH, Laranja da Fonseca, Silvia WJ. Relationship between endogenous progesterone and follicular dynamics in lact ating dairy cows with ovarian follicular cysts. Biol Reprod 2003;69:218-223.
125 Hamilton SA, Gaverick HA, Keisler DH, XU ZZ, Loos K, Younquist RS, Salfen BE. Characterization of ovarian follicular cysts and associated endocrine profiles in dairy cows. Biol Reprod 1995;53:890-898. Hampton JH, Salfen BE, Bader JF, Keisler, DH, Gaverick HA. Ovarian follicular responses to high doses of pul satile luteinizing hormone in lactating dairy cattle. J. Dairy Sci. 2003;86:1963-1969. Hanzen CH, Pieterse M, Scenczi O, Drost M. Relative accuracy of the identification of ovarian structures in the cow by ultrasounography and palpation per rectum. The Veterinary Journal 2000;159:161-170. Heersche G, Nebel RL. Measuring efficiency and accuracy of detection of estrus. J Dairy Sci 1994;77:2754-2761. Hendricks KEM. Reproductive strategies in th e postpartum dairy cow with reference to anovulation and postpartum uter ine health. MS Thesis 2004; University of Florida, Gainesville, FL. Herbison AE. Neurochemical identity of neurons expressing oe strogen and androgen receptors in sheep hypothalamus. J Reprod Fertil 1995;Suppl 49:271-283. Hooijer GA, Lubbers RBF, Ducro BJ, van Arendonk JAM, Kaal-Lansbergen LMTE, van der Lende T. Genetic parameters for cystic ovarian disease in dutch black and white dairy cattle. J Dairy Sci 2001;84:286-291 Houseknecht KL, Baile CA, Matteri RL, Spurlo ck ME. The biology of leptin: a review. J Anim Sci 1998;76:1405-1420. Hsu C-J, Hammond JM. Gonadotropins and es tradiol stimulate im munoreactive insulinlike growth factor-1 production by porcine granulosa cells in vitro Endocrinology 1987;120:198-207. Hsueh AJW, Jones PBC. Extrapituitary action of gonadotropin-releasing hormone. Endocrine Rev 1981;2:437-456. Isobe N, Nakao T, Yoshimura Y. Immunohistochemical localization on 3 hydroxysteroid dehydrogenase in the granul osa and theca interna layers of bovine cystic follicles. J Reprod Dev 2003;49:227-233. Jackson PS, Johnson CT, Furr BJ, Beattie JF. Infl uence of stage of oestrous cycle on time of oestrus following cloprostenol tr eatment in the bovine. Theriogenology 1979;12:153-168. Jansen HT, Hileman SM, Lubbers LS, Kuehl DE, Jackson GL, Lehman MN. Identification and distribution of neur oendocrine gonadotropin-releasing hormone neurons in the ewe. Biol Reprod 1997;56:655-662.
126 Jeffcoate IA, Ayliffe TR. An ultrasonographi c study of bovine cystic ovarian disease and its treatment. Vet Rec 1995;136:406-410. Johnson AD, Legates JE, Ulberg LC. Relations hip between follicular cysts and milk production in dairy cattle. J Dairy Sci 1966;49:865-868. Johnson AD, Ulberg LC. Influence of exogenous progesterone on follicular cysts in dairy cattle. J Dairy Sci 1966;50:758-761. Kaiser UB, Jakubowiak A, Steinberger A, Chin WW. Regulation of rat pituitary gonadotropin-releasing hormone receptor mRNA levels in vivo and in vitro Endocrinology 1993;133:931-934. Kalra SP, Horvath T, NaftolinF, Xu B, Pu S, Kalra PS. Intera ctive language of the hypothalamus for the gonadotropin rel easing hormone (GnRH) system. J Neuroendocrinol 1997;9:569-576. Kaltenbach CC, Dunn TG, Kiser TE, Corah LR Akbar AM, Niswender GD. Release of FSH and LH in beef heifers by synthetic gonadotropin releasing hormone. J Anim Sci 1974;38:357-362. Kaneko H, Nakanishi Y, Akagi S, Arai K, Ta ya K, Watanabe G, Sasamoto S, Hasegawa Y. Immunoneutralization of inhibin and es tradiol during the folli cular phase of the estrous cycle in cows. Biol Reprod 1995;53:931-939. Kaneko H, Todoroki J, Noguchi K, Mizoshit a K, Kubota C, Yamakuchi H. Perturbation of estradiol-feedback control of luteinizing horm one secretion by immunoneutralization induces development of follicular cysts in cattle. Biol Reprod 2002;67:1840-1845. Karsch FJ, Foster DL, Legan SJ, Ryan KD, Peter GK. Control of the preovulatory endocrine events in the ewe: interrela tionship of estradiol, progesterone, and luteinizing hormone. En docrinology 1979;105:421-426. Karsch FJ, Legan SJ, Ryan KD, Foster DL. Im portance of estradiol and progesterone in regulating LH secretion and estrous behavi or during the sheep es trous cycle. Biol Reprod 1980;32:404-413. Kasimanickam R, Cornwell JM, Nebel RL Fertility following fixed-time AI or insemination at observed estrus in Ovs ynch and Heatsynch programs in lactating dairy cows. Theriogenology 2004;63:2550-2559. Kawate N. Studies on the regulation of expre ssion of luteinizing hormone receptor in the ovary and the mechanism of follicular cyst formation in ruminants. J Reprod Dev 2004;50:1-8.
127 Kawate N, Itami T, ChoushiT, Saitoh T, Wada T, MatsuokaK, Uenaka K, Tanaka N, Yamanaka A, SakaseM, Tamada H, Inaba T, Sawade T. Improved conception in timed-artificial insemination using a progest erone-releasing intravaginal device and Ovsynch protocol in postpartum su ckled Japanese Black beef cows. Theriogenology 2004;61:399-406. Kesler DJ, Garverick HA, Caudle AB, Bier schwal CJ, Elmore RG, Youngquist RS. Clinical and endocrine responses of dair y cows with ovarian cysts to GnRH and PGF2 J Anim Sci 1978;46:719-725. Kesler DJ, Gaverick HA. Ovarian cysts in dairy cattle: Review. J Anim Sci 1982;55:1147-1159. Kim U-H, Suh G-H, Nam H-W, Kang H-G, Ki m I-H. Follicular wave emergence, luteal function and synchrony of ovulation followi ng GnRH or estradiol benzoate in a CIDR-treated, lactating Holstein cows. Theriogenology 2005;63:260-268. King ME, Kiracofe GH, Stevenson JS, Schalles RR. Effect of stage of the estrous cycle on interval to estrus after PGF2 in beef cattle. Theriogenology 1982;18:191-200. Kirk JH, Huffman EM, Lane M. Bovine cystic ovarian disease: Here ditary relationships and case study. J Am Vet Med Assoc 1982;181:474-476. Kittok RJ, Britt JH, Convey EM. Endocrine re sponse after GnRH in luteal phase cows and cows with ovarian follicula r cysts. J Anim Sci 1973;37:985-989. Kubar H, Jalakas M. Pathological changes in the reproductive organs of cows and heifers culled because of infertility. J Vet Med A 2002;49:365-372. Lauderdale JW, Seguin BE, Stellflug JN, Chenault JR, Thatcher WW, Vincent CK, Loyancano AF. Fertility of cattle following PGF2 injection. J Anim Sci 1974;38:964-967. Lehman MN, Karsch FJ. Do gonadotropin-re leasing hormone, tyrosine hydroxylase-, and -endorphin-immunoreactive neurons contain estrogen receptors? A double label immunocytochemical study in the Su ffolk ewe. Endocrinology 1993;133:887-895. Leshin LS, Rund LA, Crim JW, Kiser TE. Immunocytochemical localization of luteinizing hormone and proopiomelanocorti n neurons within the preoptic area and hypothalamus of the bovine brai n. Biol Reprod 1988;39:963-975. Levine JE. New concepts in the neuroendocrine regulation of gonadotr opin surges in rats. Biol Reprod 1997;56:293-302. Li PS, Wagner WC. Effects of hyperadrenal st ates on luteinizing hormone in cattle. Biol Reprod 1983a;29:11-23.
128 Li PS, Wagner WC. In Vivo and In Vitro st udies on the effect of adrenocorticotropic hormone or cortisol on th e pituitary response to gona dotropin releasing hormone. Biol Reprod 1983b;29:25-37. Lopez-Gatius F, Lopez-Bejar M. Reproductive performance of dairy cows with ovarian cysts after different GnRH and cl oprostenol treatments. Theriogenology 2002;58:1337-1348. Lopez-Gatius F, Santolaria P, Yaniz J, Fe nech M, Lopez-Bejar M. Risk factors for postpartum ovarian cysts and their spontaneous recovery or persis tence in lactating dairy cows. Theriogenology 2002;58:1623-1632 Lopez-Gatius F, Murugavel K, Santolaria P, Yaniz J, Lopez-Bejar M. Effects of presynchronization during th e preservice period on subse quent ovarian activity in lactating dairy cows. Th eriogenology 2003;60:545-552. Louca A, Legates JE. Production losses in dairy cattle due to da ys open. J Dairy Sci 1967;51:573-583. Lucy MC, Stevenson JS, Call EP. Control ling first service and calving interval by prostaglandin F2 gonadotropin-releasing hormone, and timed insemination. J Dairy Sci 1986;69:2186-2194. Lucy MC. Regulation of ovarian follicular growth by somatotropin and insulin-like growth factors in cattle J Dairy Sci 2000;83:1635-1647. Lucy MC. Reproductive loss in high-producing dairy cattle: Where will it end? J Dairy Sci 2001;84:1277-1293. MacLusky NJ, McEwen BS. Oestrogen modulates progestin receptor concentrations in some rat brain regions but not in others. Nature 1978;274:276-278. Macmillan KL. Oestrus synchronization with a prostaglandin analogue III. Special aspects of synchronization. N Z Vet J 1978;26:104-108. Macmillan KL, Curnow RJ, Morris GR. Oestru s synchronization with a prostaglandin analogue: II. Systems in maiden heifers. N Z Vet J 1978;26:96-98. McDonald LE. Female Reproductive System. In :McDonald, LE (Ed): Veterinary Endocrinology and Reproduction. Philadel phia: Lea & Febiger, 1980; pp. 297-299. Matteri RL, Moberg GP. Effect of cortis ol or adrenocorticotropin on release of luteinizing hormone induced by luteinizing hormone releasing hormone in the dairy heifer. J Endocrinol 1982;92:141-146.
129 Mazerbourg S, Jurgen Z, Bar RS, Brigstock DR, Monget P. Insulin-like growth factor (IGF)binding protein-4 pr oteolytic degradation in bovine, equine, and porcine preovulatory follicles: regul ation by IGFs and heparin-binding domain-containing peptides. Biol Reprod 2000;63:390-400. Melendez P, Bartolome J, Archbald LF, Donovan A. The association between lameness, ovarian cysts and fertility in lactat ing dairy cows. Theriogenology 2003;59:927937. Millar RP, Lu Z-L, Pawson AJ, Flanagan CA, Morgan K, Maudsley SR. Gonadotropinreleasing hormone receptors. Endocrine Rev 2004;25:235-275. Mihm M, Austin EJ, Good TEM, Ireland JLH, Knight PG, Roche JF, Ireland JJ. Identification of potential intrafollicular f actors involved in selection of dominant follicles in heifers. Biol Reprod 2000;63:811-819. Moberg GP. Site of action of endotoxins on hypothalamic-pituitaryadr enal axis. Am J of Physio 1971;220:397-400. Moenter SM, Karsch FJ, Lehman MN. Fos expression during th e estradiol-induced gonadotropin-releasing hormone (GnRH) surg e of the ewe: induction in GnRH and other neurons. Endoc rinology 1993;133:896-903. Moenter SM, Caraty A, Locatelli A, Kars ch FJ. Pattern of gonadotropin-releasing hormone (GnRH) secretion leading up to ovulation in the ewe: existence of a preovulatory GnRH surge. Endocrinology 1991;129:1175-1182. Momcilovic D, Archbald LF, Walters A, Tran T, Kelbert D, Risco C, Thatcher WW. Reproductive performance of lactating dairy cows treated with gonadotropinreleasing hormone (GnRH) and/or prostaglandin F2 for synchronization of estrus and ovulation. Theriogenology 1998;50:1131-1139. Moore DA, Overton MW, Chebel RC, Trus cott ML, BonDurrant RH. Evaluation of factors that affect embryonic loss in dairy cattle. J Am Vet Med Assoc 2005;266:1112-1118. Moreira F, Risco C, Rires FA, Ambrose JD Drost M, DeLorenzo M, Thatcher WW. Effect of body condition on reproductive efficiency of lactating dairy cows receiving a timed inseminati on. Theriogenology 2000a;53:1305-1319. Moreira F, de la Sota RL, Diaz T, Thatcher WW Effect of day of th e estrous cycle at the initiation of a timed artificial inseminati on protocol on reproductive responses in dairy heifers. J Anim Sci 2000b;78:1568-1576. Morrow DA, Roberts SJ, McEntee K, Gray HG. Postpartum ovarian activity and uterine involution in dairy cattle. J Am Vet Med Assoc 1966;149:1596-1609.
130 Moss GE, Crowder ME, Nett TM. GnRH-receptor interaction VI. Effect of progesterone and estradiol on hypophyseal receptors for GnRH, and serum and hypophyseal concentrations of gonadotropins in ovariectomized ewes. Biol Reprod 1981;25:938-944 Nanda AS, Ward WR, Dobson H. Lack of LH response to oest radiol treatment in cows with cystic ovarian disease and effect of progesterone treatment or manual rupture. Res Vet Sci 1991;51:180-184. National Research Council. Nutrient requirements of dairy cattle, 7th Edition. Washington, D.C.: National Academy Press, 2001. Nebel RL, Whittier WD, Cassell BG, Britt JH. Comparison of on-farm and laboratory milk progesterone assays for identifying erro rs in detection of estrus and diagnosis of pregnancy. J Dairy Sci 1987;70:1471-1476. Nett TM, Crowder ME, Moss GE, Duello TM. GnRH-receptor interaction. V. Downregulation of pituitary receptors for GnRH in ovariectomized ewes by infusion of homologous hormone. Biol Repord 1981;24:1145-1155. Odore R, Re G, Badino P, Donn A, Vigo D, Biolatti B, Girardi C. Modification of receptor concentrations for adrenaline, steroid hormones, prostaglandin F2 and gonadotropins in hypophysis and ovary of da iry cows with ovarian cysts. Pharm Res 1999;39:297-304. Ojeda SR, McCann SM. The ante rior pituitary and hypothalamu s. In: Griffin JE, Ojeda SR (Eds): Textbook of Endocrine Physiology, 4th Edition. New York, Oxford University Press, 2000; pp. 128-162. Olds D, Cooper T, Thrift FA. Effect of days open on economic aspects of current lactation. J Dairy Sci 1979;62:1167-1170. Ozturk M, Smith RF, Dobson H. Effect of prolonged exposure to oestradiol on subsequent LH secretion in ewes. J Reprod Fertil 1998;114:1-9. Padmanabhan V, Keech C, Convey EM. Cortis ol inhibits and adrenocorticotropin has no effect on luteinizing hormonereleasing hormoneinduced release of luteinizing hormone from bovine pituitary cells in vitro Endocrinology 1983;112:1782-1787. Padmanabhan V, Brown MB, Dahl GE, Evan s NP, Karsch FJ, Mauger DT, Neill JD, Van Cleeff J. Neuroendocrine control of follic le-stimulating hormone (FSH) secretion: III. Is there a gonadotropin-releasing hor mone-independent component of episodic FSH secretion in ovariectomized a nd luteal phase ewes? Endocrinology 2003;144:1380-1392. Pankowski JW, Galton DM, Erb HN, Guard CL, Grohn YT. Use of prostaglandin F2 as a postpartum reproductive management tool for lactating dairy cows. J Dairy Sci 1995;78:1477-1488.
131 Perry GA, Smith MF, Lucy MC, Green JA, Parks TE, MacNeil MD, Roberts AJ, Geary TW. Relationship between follicle size at insemination and pregnancy success. Proc National Acad Sci 2005;102:5268-5273 Peter AT, Bosu WKT, DeDecker RJ. Supre ssion of preovulatory luteinizing hormone surges in heifers after in trauterine infusions of Escherichia coli endotoxin. Am J Vet Res 1989;50:368-373. Peter AT. Infertility due to abnormalities of the ovaries. In: Youngquist RS (Ed.), Current Therapy in Large Animal Theri ogenology. Philadelphia: WB Saunders Company, 1997. pp. 349-354. Peters MW, Pursley JR. Fertility of lacta ting dairy cows treated with Ovsynch after presynchronization in jections of PGF2 and GnRH. J Dairy Sci 2002;85:2403-2406. Peters MW, Pursley JR. Timing of final GnRH of the Ovsynch protocol affects ovulatory follicle size, subsequent luteal function, and fertility in dairy cows. Theriogenology 2003;60:1197-1204. Peterson SL, Ottem EN, Carpenter CD. Direct and indirect regul ation of gonadotropinreleasing hormone neurons by estr adiol. Biol. Reprod 2003;69:1771-1778. Pierson RA, Ginther OJ. Untrasonography of the bovine ovary. Theriogenology 1984;21:495-504. Pierson RA, Ginther OJ. Ultrasonographic imagi ng of the ovaries and uterus in cattle. Theriogenology 1988;29:21-37. Pursley JR, Mee MO, Wiltbank MC. Synchroni zation of ovulation in dairy cows using PGF2 and GnRH. Theriogenology 1995;44:915-923. Pursley JR, Kosorak MR, Wiltbank MC. Repr oductive management of lactating dairy cows using synchronization of ovulation. J Dairy Sci 1997;80:301-306. Pursley JR, Silcox RW, Wiltbanl MC. Effect of time of artificial insemination on pregnancy rates, calving rates, pr egnancy loss, and gender ratio after synchronization of ovulation in lactat ing dairy cows. J Dairy Sci 1998;81:21392144. Rajakoski E. The ovarian follicular system in sexually mature heifers with special reference to seasonal, cyclical, and left right variations. Ac ta Endocrinol Suppl. 1960;52:6. Ramirez-Godinez JA, Kiracofe GH, McKee RM, Schalles RR, Kittok RJ. Reducing the incidence of short estrous cycles in beef cows with norgestomet. Theriogenology 1981;15:613-623.
132 Rani CS, Salhanick AR, Armstrong DT. Fo llicle-stimulating hormone induction of luteinizing hormone receptor in cultured ra t granulosa cells: an examination of the need for steroids in the inducti on process. Endocrinology 1981;108:1379-1385. Reeves JJ, Arimura A, Schally AV. Changes in pituitary responsiveness to luteinizing hormone-releasing hormone (LH-RH) in anes trous ewes pretreated with estradiol benzoate. Biol Repord 1971;4:88-92. Refsal KR, Jarrin-Maldonadl JH, Nachreiner RF Endocrine profiles in cows with ovarian cysts experimentally induced by trea tment with exogenous estradiol or adrenocorticotropic hormone Theriogenology 1987;28:871-889. Refsal KR, Jarrin-Maldonadl JH, Nachreiner RF. Basal and estradio l induced release of gonadotropin in dairy cows with natura lly occurring cysts. Theriogenology 1988;30:679-693. Ribadu AY, Ward WR, Dobson H. Comparative ev aluation of ovarian st ructures in cattle by palpation per rectum, ultrasonography, a nd plasma progesterone concentration. Vet Rec 1994;135:453-457. Ribadu AY, Nakada K, Tanaka Y, Moriyoshi M, Zhang WC, Nakao T. Lack of LH response to exogenous estradiol in heifer s with ACTH-induced ovarian follicular cysts. J Vet Med Sci. 1999;61:979-981. Richards JS. Maturation of ova rian follicles: actions and in teractions of pituitary and ovarian hormones on follicu lar cell differentiation. P hysiological Rev 1980;60:5789. Rivera GM, Fortune JE. Developm ent of codominant follicles in cattle is associated with a follicle-stimulating hormone-dependent insulin-like growth factor binding protein-4 protease. Bi ol Reprod 2001;65:111-118. Rivera GM, Fortune JE. Proteolysis of insulin -like growth factor bi nding proteins -4 and -5 in bovine follicular fluid: implicati ons for ovarian follicular selection and dominance. Endocrinology 2003;144:2977-2987. Roberts, SJ. Veterinary Ob stetrics and Genital Diseas es: Theriogenology. Edward Brothers Inc., Ann Arbor Michigan, 1971, pp. 422-423. Robinson RS, Hammond AJ, Hunter MG, Ma nn GE. The induction of a delayed postovulatory progesterone rise in dairy cows: a novel model. Domest Anim Endocrinol 2005;28:285-295. Roche JF. Effect of short-term progesterone treatment on oestrous response and fertility in heifers. J Repr od Fertil 1974;40:433-440. Roche JF. Control of time of ovulation in heifers treated with progesterone and gonadotropin-releasing hormone. J Reprod Fertil 1975;43:471-477.
133 Roche JF. Retention rate in cows and heifer s of intravaginal silastic coils impregnated with progesterone. J Reprod Fertil 1976;46:253-255. Rodriguez TR, Fields MJ, Burns WC, Franke DE, Hentges JF, Thatcher WW. Breeding at a predetermined time in the bovine following PGF2 + GnRH. J Anim Sci 1975;40:188. Rounsaville TR, Oltenacu PA, Milligan RA, Foote RH. Effects of heat detection, conception rate, and culling policy on repr oductive performance in dairy herds. J Dairy Sci 1979;62:1435-1442. Rowson LEA, Tervit R, Brand A. The use of prostaglandins for synchronization of oestrus in cattle. J Reprod Fertil 1972;29:145. Roy SK, Greenwald GS. Hormonal requirements for the gr owth and differentiation of hamster preantral follicles in long-t erm culture. J Reprod Fertil 1989;87:103-114. Rutter LM, Carruthers TD, Manns JG. The postp artum induced corpus luteum: functional differences from that of cycling cows and the effects of progesterone pretreatment. Biol Reprod 1985;33:560-568. Ryan DP, Snijders S, Yaakub H, OFarrell KJ An evaluation of es trus synchronization programs in reproductive management of dairy herds. J Anim Sci 1995;73:36873695. Ryan DP, Galvin JA, OFarrell KJ. Comparis on of oestrous synchr onization regimens for lactating dairy cows. An im Reprod Sci 1999;56:152-168. Santos JEP, Bartolome JA, Cerri RLA, Juch em SO, Hernandez O, Trigg T, Thatcher WW. Effect of a deslorelin implant in a timed artificial insemination protocol on follicle development, luteal function and reproductive performance of lactating dairy cows. Theriogenology 2004a;61:421-435. Santos JEP, Juchem SO, Cerri RLA, Galvao KN, Chebel RC, Thatcher WW, Dei CS, Bilby CR. Effect of bST and reprodu ctive management on preproductive performance of Holstein dair y cows. J Dairy Sci 2004b;87:868-881. Sarkar DK, Fink G. Effects of gonadal ster oids on output of luteinizing hormone releasing factor into pituitary stalk blood in the female rat. J Endocrinol 1979;80:303-313. Sartori R, Fricke PM, Ferreira JCP, Ginthe r OJ, Wiltbank MC. Follicular deviation and acquisition of ovulatory capacity in bovi ne antral follicles. Biol Reprod 2001;65:1403-1409. Savio JD, Boland MP, Hynes N, Mattiacci MR, Roche JF. Will the first dominant follicle of the estrous cycle of heifers ovul ate following luteolysis of Day 7. Theriogenology 1990;33:677-686.
134 Savio JD, Thatcher WW, Morris GR, Entwistl e K, Drost M, Mattiacci MR. Effects of induction of low plasma progesterone concen trations with a pr ogesterone-releasing intravaginal device on follicular turnover and fertility in cattle. J Reprod Fertil 1993;98:77-84 Schmitt EJ-P, Diaz T, Barros CM, de la Sota RL, Drost M, Fredriksson EW, Staples CR, Thorner R, Thatcher WW. Differential respons e of the luteal phase and fertility in cattle following ovulation of the first-wave follicle with human chorionic gonadotropin or an agonist of gonadotr opin-releasing hormone. J Amin Sci 1996;74;1074-1083. Scott CJ, Tilbrook AJ, Rawson JA, Clarke IJ. Gonadal steroid receptors in the regulation of GnRH secretion in farm animals. Anim Reprod Sci 2000;60-61:313-326. Shen ES, Meade EH, Perez MC, Deecher DC, Ne gro-Villar A, Lopez FJ. Expression of functional estrogen receptors and ga lanin messenger ribonucleic acid in immortalized luteinizing hormone-releas ing hormone neurons: estrogenic control of galanin gene expression. Endocrinology 1998;139:939-948. Shivers BD, Harlan RE, Morrell JI, Pfaff DW. Ab sence of oestradiol concentration in cell nuclei of LHRH-immunoreactive neurons. Nature 1983;304:345-347. Shughrue PJ, Lane MV, Merchenthaler I. Re gulation of progesterone receptor messenger ribonucleic acid in the rat medial preoptic nucleus by estrogenic and antiestrogenic compounds: an in situ hybridization study. E ndocrinology 1997;138:5476-5484. Silvia WJ, Halter TB, Nugent AM, Laranja da Fonseca LF. Ovarian follicular cysts in dairy cows: an abnormality in folliculogenesis. Domest Anim Endocrinol 2002;23:167-177. Simerly RB, Young BJ. Regulation of estrogen receptor messenger ribonucleic acid in rat hypothalamus by sex steroid hormones. Mol Endocrinol 1991;5:424-432. Slenning BD. Quantative tools for production-or ientated veterinarians. In: Radostits OM (Ed.), Herd Health Food An imal Production Medicine 3rd Edition. WB Saunders Company, New York, 2001. pp 47-106. Smith RD, Pomerantz AJ, Beal WE, McCann JP Pilbeam TE, Hansel W. Insemination of Holstein heifers at a preset time af ter estrous cycle synchronization using progesterone and prostaglandi n. J Amin Sci 1984;58:792-800. Smith MW, Stevenson JS. Fate of the dom inant follicle, embryonal survival, and pregnancy rates in dairy cattle treated with prostaglandin F2 and progestins in the absence or presence of a functional corpus luteum. J Amin Sci 1995;73:3743-3751. Smith MJ, Wise PM. Neurotensin gene expr ession increases during proestrus in the rostral medial preoptic nucleus: pot ential for direct communication with gonadotropin-releasing hormone ne urons. Endocrinology 2001;142:3006-3013.
135 Spicer LJ, Alpizar E, Echternkamp SE. Effect s of insulin, insulin-like growth factor I, and gonadotropins on bovine granulosa cell proliferation, proges terone production, estradiol production, and (or) insuli n-like growth factor I production in vitro J Anim Sci 1993;71:1232-1241. Spicer LJ, Echternkamp SE. The ovarian insuli n and insulin-like growth factor system with an emphasis on domestic animal s. Domest Anim Endocrinol 1995;12:223245. Spicer LJ, Alvarez P, Prado TM, Morgan GL, Hamilton TD. Effects of intraovarian infusion of insulin-like growth factor on ovarian follicular function in cattle. Domest Anim Endocrinol 2000;18:265-278. Sprecher DL, Nebel RL, Whittier WD. Predictive value of palpation per rectum versus milk and serum progesterone levels for the diagnosis of bovine follicular and luteal cysts. Theriogenology 1988;30:701-710. Staples CR, Thatcher WW. Relationship be tween ovarian activity and energy status during the early postpartum period of high producing dairy cows. J Dairy Sci 1990;73:938-947. Stevenson JS, Lucy MC, Call EP. Failure of timed inseminations and associated luteal function in dairy cattle after tw o injections of prostaglandin F2 Theriogenology 1987;28:937-946. Stevenson JS, Kobayashi Y, Shipka MP, Rauchholz KC. Altering conception of dairy cattle by gonadotropin-releasing horm one preceding luteolysis induced by prostaglandin F2 J Dairy Sci 1996;79:402-410. Stevenson JS, Koybayshi Y, Thompson KE. Reproductive performance of dairy cows in various breeding systems including Ovs ynch and combinations of gonadotropinreleasing hormone and prostaglandin F2 J Dairy Sci 1999;82:506-515. Stevenson JS. Clinical reproductive physiol ogy of the cow. In: Youngquist RS (Ed.), Current Therapy in Large Animal Theri ogenology. Philadelphia: WB Saunders Company, 1997. pp.257-267. Stewart RE, Spicer LJ, Hamilton TD, Keefer BE Effects of insulinlike growth factor 1 and insulin on proliferation and on ba sal and luteinizi ng hormone-induced steroidogenesis of bovine th ecal cells: involvement of glucose and receptors for insulin-like growth fact or 1 and luteinizing ho rmone. J Anim Sci 1995;73:37193731. Stoebel DP, Moberg GP. Repeated acute stre ss during the follicular phase and luteinizing hormone surge of dairy heifers. J Dairy Sci 1982;65:92-96.
136 Stumpf TT, Roberson MS, Wolfe MW, Hamernik DL, Kittok RJ, Kinder JE. Progesterone, 17 -estradiol and opioid neurope ptides modulate pattern of luteinizing hormone in circulation of the cow. Biol Reprod 1993;49:1096-1101. Sullivan SD, Moenter SM. Aminobutyric acid neurons integrate and rapidly transmit permissive and inhibitory metabolic cues to gonadotropin-releasing hormone neurons. Endocrinology 2004;145:1194-1202. Sullivan SD, Moenter SM. GABAergic integration of progesterone and androgen feedback to gonadotropin-releasing hor mone neurons. Biol Reprod 2005;72:33-41. Tanabe TY, Hann RC. Synchroni zed estrus and subsequent conception in dairy heifers treated with prostaglandin F2 I. Influence of stage of cycle at treatment. J Anim Sci 1984;58:805-811. Tasende C, Meikle A., Rodriguez-Pinon M, Forsberg M, Garofalo EG. Estrogen and progesterone receptor content in the pitu itary gland and uterus of progesteroneprimed and gonadotropin releasing hormonetreated anestrous ewes. Theriogenology 2002;57:1719-1731. Tenhagen B-A, Drillich M, Surholt R, Heuw iesser W. Comparison of timed AI after synchronized ovulation to AI at estrus: reproductive and economic considerations. J Dairy Sci 2004;87:85-94. Thurston LM, Jonas KC, Abayasekara RE, Micheal A. Ovarian modulators of 11 hydroxysteroid dehydrogenase activity in fol licular fluid from bovine and porcine large antral follicles and spontaneous ovarian cysts. Biol Reprod 2003;68:21572163 Todoroki J, Yamakuchi H, Mizoshita K, Kubot a N, Tabara N, Noguchi J, Kikuchi K, Watanabe G, Taya K, Kaneko H. Restor ing ovulation in beef donor cows with ovarian cysts by progesterone releasi ng intravaginal silastic devices. Theriogenology 2001;55:1919-1932. Tsai H-W, Legan SJ. Loss of luteinizing horm one surges induced by chronic estradiol is associated with decreased activation of gonadotropin-releasing hormone neurons. Biol Reprod 2002;66:1104-1110. Turzillo AM, Campion CE, Clay CM, Nett TM. Regulation of gonadotropin-releasing hormone (GnRH) receptor messenger ribonuc leic acid and GnRH receptors during the early preovulatory period in the ewe. Endocrinology 1994;135:1353-1358. Twagiramungu H, Guilbault LA, Proulx J, Vill eneuve P, Dufour JJ. Influence of an agonist of gonadotropin-releasing hormone (Burselin) on estrus synchronization and fertility in beef cows. J Anim Sci 1992;70:1904-1910.
137 Van Eerdenburg FJCM, Karthaus D, Taverne MA M, Merics I, Szenci O. The relationship between estrous behavioral score and time of ovulation in dairy cattle. J Dairy Sci 2002;85:1150-1156. Vasconcelos JLM, Silcox RW, Rosa GJM, Pursley JR, Wiltbank MC. Synchronization rate, size of the ovulatory follicle, and pr egnancy rate after synchronization of ovulation beginning on different days of the estrous cycle in lactating dairy cows. Theriogenology 1999;52:1067-1078. Vizcarra JA, Wettemann RP, Braden TD, Turz illo AM, Nett TM. Effect of gonadotropinreleasing hormone (GnRH) pulse frequenc y on serum and pituitary concentrations of luteinizing hormone and follicle st imulating hormone, GnRH receptors, and messenger ribonucleic acid for gonadotr opin subunits in cows. Endocrinology 1997;138:594-601. Webb R, England BG, Fitzpatrick KE. Contro l of the preovulatory gonadotropin surge in the ewe. Endocri nology 1981;108:1178-1185. Wehrman ME, Roberson MS, Cupp AS, Kojima FN, Stumpf TT, We rth LA, Wolfe MW, Kittok RJ, Kinder JE. Increasing exogenous progesterone during s ynchronization of estrus decrease endogenous 17 -estradiol and increases conception in cows. Biol Repord 1993;49:214-220. White ME, Erb H. Treatment if ovarian cysts in dairy cattle: a decision analysis. Cornell Vet 1980;70:247-257. White ME, Erb H. Optimum postpartum interv al for screening dairy cows for ovarian cysts by rectal palpation. Cornell Vet 1982;72:137-141. Whitmore HL, Tyler WJ, Casida LE. Incidence of cystic ovaries in Holstein-Fresian cows. J Am Vet Med Assoc 1974;165:693-694. Whitmore HL, Hurtgen JP, Mather EC, Seguin BE. Clinical response of dairy cattle with ovarian cysts to single or repeated tr eatments of gonadotropin-releasing hormone. J Am Vet Med Assoc 1979;174:1113-1115. Wiltbank JN, Kasson CW. Synchronization of estrus in cattle with an oral progestatational agent and an injecti on of an estrogen. J Anim Sci 1968;27:113116. Wiltbank MC, Gumen A, Sartori R. Phys iological classification of anovulatory conditions in cattle. Theriogenology 2002;57:21-52. Wolfenson D, Thatcher WW, Savio JD, Badi nga L, Lucy MC. The effect of a GnRH analogue on the dynamics of follicular deve lopment and synchronization of estrus in lactating cyclic dairy cows. Theriogenology 1994;42:633-644.
138 Xu Z, Garverick HA, Smith GW, Smith MF Hamilton SA, Youngquist RS. Expression of follicle-stimulating hormone and luteinizing hormone receptor messenger ribonucleic acids in bovine follicles during the first follicular wave. Biol Reprod 1995;53:951-957. Xu ZZ, Burton LJ, Macmillan KL. Reproductive performance of lactating dairy cows following oestrous synchronization w ith progesterone, oestradiol and prostaglandin. N Z Vet J 1996;44:99-104. Xu ZZ, Burton LJ, Macmillan KL. Reproductive performance of lactating dairy cows following estrus synchronization regimens with PGF2 and progesterone. Theriogenology 1997;47:687-701. Xu M, Urban JH, Hill JW, Levine JE. Regulation of hypothalamic neuropeptide Y Y1 receptor gene expression during the estrous cycle: role of progesterone receptors. Endocrinology 2000;141:3319-3327. Yang MY, Rajamahendran R. Effects of gonadot ropins and insulin-like growth factors-I and II: in vitro steroid production by bovine gr anulosa cells. Can J An Sci 1998;78:587-597. Yoshioka K, Iwamura S, Kamomae H. Ultras onic observation on the turnover of ovarian follicular cysts and associated changes of plasma LH, FSH, progesterone and oestradiol-17 in cows. Res Vet Sci 1996;61:240-244. Youngquist RS. Cystic follicular degeneration in the cow. In: Morrow DA (Ed.), Current Therapy in Theriogenology 2. Philade lphia: WB Saunders Company, 1986. pp.243246. Zaied AA, Gaverick HA, Kesler DJ, Bier schwal CJ, Elmore RG, Youngquist RS. Luteinizing hormone response to estradiol benzoate in postpartum cows and cows with ovarian cysts. Th eriogenology 1981;16:349358. Zieba DA, Amstalden M, Morton S, Maciel MN, Keisler DH, Williams GL. Regulatory roles of leptin at the hypothalamic-h ypophyseal axis before and after sexual maturation in cattle. Bi ol Reprod 2004;71:804-812. Zemjanis R. Diagnostic and therapeutic tec hniques in animal reproduction. Baltimore: Williams and Wilkins Company, 1970. pp.29-46. Zulu VC, Nakao T, Yamada K, Moriyoshi M, Nakada K, Sawamukai Y. Clinical response of ovarian cysts in dairy cows after PRID treatment. J Vet Med Sci 2003;65:57-62. Zwald NR, Weigel KA, Chang YM, Welper RD Clay JS. Genetic selection for health traits using producer recorded data. I. In cidence rates, heritability estimates and sire breeding values. J Da iry Sci 2004a;87:4287-4294.
139 Zwald NR, Weigel KA, Chang YM, Welper RD Clay JS. Genetic selection for health traits using producer recorded data. II. Genetic correlations, disease probabilities and relationships with existing traits. J Dairy Sci 2004b;87:4295-4302
140 BIOGRAPHICAL SKETCH Mary Bronwyn Crane was born on August 3, 1978, in Prince Edward Island, Canada. She grew up in the small town of Kensington, P.E.I., where her family had a small farm and raised purebred beef cattle. She had a strong inclination towards animals and enjoyed working with her father who wa s a rural veterinarian. After completing high school in Kensington, she attended Acadia Univer sity in Wolfville, Nova Scotia, enrolled in a Bachelor of Science program. Within two years she completed the pre-veterinary curriculum at Acadia University, then was accep ted at the Atlantic Veterinary College in Charlottetown, P.E.I. Upon graduation from veterinary college in 2002, she moved to Gainesville, Florida, where she joined a small animal practice as an associate veterinarian. After one year of small anim al practice in 2003 she decided to pursue her long time interest in food animal practice a nd applied to the Farm Animal Reproduction and Medicine Service Internship program at the College of Veterinary Medicine, University of Florida, and was accepted. Afte r completing the one year internship in 2004, she enrolled in the graduate program at the Department of Large Animal Clinical Sciences, College of Veterinary Medicine, Univ ersity of Florida, to obtain the degree of Master of Science.