,_PEER REVIEWED

Dinoprost Versus Cloprostenol: Does Route of InjectionModulate theirEfficacy in Dairy Cattle?

Roger Martineau, DVM, 253 chemin Bulwer, Martinuille, Québec, Canada JOB2AO

Abstract

This factorial study was conducted to determinethe reproductive performance of402 dairy cattle treatedwith either dinoprost (25 mg) or cloprostenol (500 llg),intravenously (lV) or intramuscularly (lM), for synchro-nization of estrus in a randomized clinical trial. Breed-ing, conception and pregnancy rates were analyzed withmultivariate logistic regression. The model included thetreatments (2 types of prostaglandins and 2 routes ofinjection) as weIl as eight clinically-relevant covariables.Interactions between treatments were not significantand comparisons were made on the main effect. Com-parisons between types of prostaglandins (PGF2a)showed similar breeding rates (82.6 vs 83.0%; P=0.93),but lower conception rate (38.6 vs 46.6%; P=0.13) andpregnancy rate (31.4 vs 39.2%; P=0.14) for dinoprostversus cloprostenol. When comparing routes ofinjection,a similar breedingrate (84.1vs 81.4%;P=0.45),but lowerconception rate (38.7 vs 46.7%; P=0.13) andpregnancyrate (32.5 vs 38.1%; P=0.23) were noted forIM versusIVinjections. The statistical power ofthe study was 19%,which allowed detection (P

vaches inséminées quatre jours ou plus après l'injectionde prostaglandine plutôt qu'inséminées avant (69.4%versus 51.9%; P = 0.08). On conclut que le type deprostaglandine utilisé de même que la voie d'injectionpeuvent influencer le taux de conception, et parconséquent le taux de gestation, mais que le pouvoirstatistique de l'étude était trop faible pour détecter desdifférences significatives. On conclut aussi quel'intervalle de temps entre l'injection de prostaglandineet la saillie de même que le type de prostaglandine et lavoie d'injection ont une influence sur le rapport des sexesdes veaux.

Introduction

ln Canada, two types of prostaglandins (PGF2a)are popular: dinoprost, a tromethamine salt (THAM)ofthe natural PGF2a, and cloprostenol, a synthetic ana-logue. Both products can be injected intramuscularly(IM) or intravenously (IV), but the lM route is recom-mended by manufacturers and is more desirable from astand point of commercial application than the IV injec-tion.6.l2 Cloprostenol has a longer biological half-life andis a much more potent luteolytic agent than dinoprostsince it is not degraded by 15-hydroxydehydrogenaseand 13,14-reductase.1 The IV injected PGF2a is metabo-lized during the first few passages through the lungs,resulting in a shorter peripheral exposure than the lMinjected PGF2a, which is released more slowlyfrom theinjection site. 13Verylimited research has been conductedto compare the reproductive efficiency obtained withthese two PGF2a products and no clinical trial adressingthe route of injection issue was found in the literature.

A recent review ofthe literature pertaining to theeffect oftime ofinsemination on sex ratio indicates thattreatments used for synchronization of estrus or ovula-tion in caUle may influence the sex ratio. 15Yet no infor-mation was found on the effects of type of PGF2a usedand route of injection on the sex ratio of calves.

Unti11992, veterinarians from the Coaticook Vet-erinary Clinic (Coaticook, Québec, Canada) were exclu-sively using cloprostenol IV. ln 1992, they switched fromcloprostenol IV to dinoprost lM. Many dairy producerswere unsatisfied with the new PGF2a and the new routeofinjection. This prompted the initiation of a field studyto compare the reproductive efficiencyofanimaIs treatedwith each of the two types of PGF2a given lM or IV. Itwas hypothesized that both treatments wouldyield simi-lar results and have no effect on the sex ratio of calves.

Material and Methods

DesignLactating dairy cows and heifers past their volun-

tary waiting period and having a corpus luteum (CL),

11

,;,.as determined by rectal palpation, were assigned ran-~'domly to a2 x 2 factorial design that included a··dinoprosta(25 mg or 5 ml) and a cloprostenolb (500 Jlgor 2 ml) group injected either IMc (gluteal muscle orcroup) or IV (coccygeal vein). AlI records were main-tained in DSA softwared (Dossier de Santé Animale)from ASTLQ.e SampIe size was not calculated beforethe experiment, but a goal of 100 animaIs in each treat-ment group was considered realistic.

Animal SelectionThe study was conducted from May 1, 1993to June

1, 1994 on dairy caUle from 152 farms participating ina monthly herd health program. The final data set in-cluded 402 valid cases after 10 animaIs were culled sub-sequent to synchronization of estrus, butprior topregnancy diagnosis. Owner permission was requestedbefore including an animal in the study. Most partici-pants (n=69) had a tie staIl operation and cows wereallowed to graze on pasture in the summer. The aver-age herd had 54 (range 26-123) Holstein cows with a305-day mature milk production of approximately17,600 lb (8,000 kg). Artificial insemination (AI) wasperformed following an observed estrus, although sorneproducers occasionally used fixed-time breeding 3 and4 days post-treatment. Inseminations were done by ateam of experienced AI technicians. Pregnancy was de-termined via rectal palpation 35 to 60 days followingAI. Sex of calves was not determined prior to calving.

ProcedureAtechnical assistant prepared 480 sealed and num-

bered envelopes containing two syringes (2 ml and 5ml), onefilled with one of the two commercial PGF2aproducts under study (dinoprost or cloprostenol) and theother with sterile physiological saline. Each envelopecontained a note indicating the route of injection to beused and was randomly allocated to a participating ani-mal. Both syringes were injected by the same route tothe same animal in order to avoid selection bias and toblind both the producer and the veterinarian injectingthe PGF2.a

Data Collection and DefinitionsThe followinginformation was gathered at the farm

and included in the database: date ofinjection, herd andanimal identification, and the name of the veterinarianinjecting the PGF2a product. The following informationwas also included in the database: birth date, lactationnumber, calving date, treatment, days in milk (DIM),breeding dates, cystic ovarian disease (COD) diagnosis,pregnancy examination date and pregnancy outcome.Since heat signs were not routinely recorded by aIl pro-ducers, they were not included in the analysis. Only thebreedings made within 7 days after the PGF2a injection

THE BOVINE PRACTITIONER-VOL. 37, NO. 1

,.,

were considered in the study. When multiple insemina-tions were performed during the same estrous cycle,onlythe first breeding was considered in the analysis (15cases overall, 7 cases in herd 11; Table 3).

Three relevant fertility rates as well as the malecalf rate were considered as dependent variables. Breed-ing rate wasdefined as the number of cattle insemi-nated within 7 days post-injection divided by the numberofcaUle assigned. Conception and pregnancy rates werederived from pregnancy status at first transrectal ex-amination following the induced estrus ..Pregnancy losswas not considered. Conception rate was defmed as thenumber of pregnant caUle divided by the number ofin-seminated caUle, and pregnancy rate was the numberof pregnant cattle divided by the number of cattle as-signed. The male calfrate was the number ofmale calvesdivided by the number of calves born with recorded gen-der. Each twin counted as a single calf for the male calfrate analysis.

Statistical AnalysisChi-Square tests of independence and logistic re-

gressionanalysis were conducted with Statistix.c Sig-nificance was declared at P

Table 2. Descriptive data on the 4 treatment groups.

Dinoprost Cloprostenol

Intramuscular Intravenous Intramuscular Intravenous

Heifers, n 17 18 18 17Cows,n 85 81 83 83Lactation number-mean (sd) 2.14 (1.88) 2.37 (1.95) 2.36 (2.07) 2.45 (2.01)-range 0-9 0-8 0-8 0-9

Days in milk,bdays-mean (sd) 142 (72.2) 134 (69.4) 133 (58.4) 122 (52.1)-range 59-426 55-473 60.:..-374 52-283

Ageofheifers,bdays-mean (sd) 591(114.7) 624 (123.1) 617 (109.4) 608 (80.0)-range 423-847 447-928 430-940 459-766

Ale number-mean (sd) 0.88 (1.14) 0.87 (1.38) 0.76 (1.06) 0.64 (1.05)-range 0-5 0-7 0-6 0-6

pa

0.70

0.24

0.83

0.42

ap-valuefromOne-WayAnovabAtPGF2"treatmenteArtificialinsemination beforePGF2"treatment for cowsand heifers

tation when PGF2a was injected) needed to be catego-rized because their association with outcomes was notlinear.8 The breeding rate was significantly associatedwith the number of caUle injected duringthe same visit(grouping ofinjections), lactational incidence ofCOD, theveterinarian who injected the PGF2a and the herd (Table3). The conception and the pregnancy rates were signifi-cantly associated with lactation number, grouping ofin-jections and stage of lactation. The pregnancy rate wasalso significantly associated with lactational incidence ofCOD. The male calfrate was not significantly associatedwith any of the listed covariables. Mean days from injec-tion to Al was shorter in heifers compared to lactatingCOVIS (3.25 days, SE=0.086 vs 3.45 days, SE=0.047;P=0.05) and not different between types of PGF2a (3.42days, SE=0.061 vs 3.40 days, SE=0.056; P=0.86) and be-tween routes of injection (3.43 days, SE=0.062 vs 3.40days, SE=0.054; P=0.65).

The significance levels for the interaction betweentreatments (PGF2a x ROUTE) were 0.37, 0.99 and 0.64for the breeding rate, the conception rate and the preg-nancy rate, respectively. No interactions between thelisted covariables and between treatments were retainedin the multivariate logistic regression model. Estimatedodds ratios and 95%confidence intervals (CI) for the com-parison ofcloprostenol against dinoprost were 1.03 (0.59,1.82), 1.42 (0.88, 2.30) and 1.41 (0.90, 2.20), respectivelyfor breeding, conception and pregnancy rates. For thecomparison of IV against lM injection, they were 0.83(0.45, 1.50), 1.39 (0.83,2.32) and 1.28 (0.80, 2.05), respec-tively, for breeding, conception and pregnancy rates.

13

Fertility odds ratios were converted to rates, whichare first presented for simple effects of each treatmentgroup (Table 4) and for the main effect of treatments(types of PGF2a and routes of injection), after poolingdata (Table 5). Fertility rates were not statiscally dif-ferent between treatment groups, between types ofPGF2a used and between routes of injection. Concep-tion rate was nearly 10%higher for each type ofPGF2awhen injected IV rather than lM (+9.9%, P=0.21 fordinoprost IV and +9.7%,P=0.19 for cloprostenol IV;Table4). When using pooled data, conception rate was also8.0% (P=0.13) higher for cloprostenol than dinoprost andwas 8.0% (P=0.13) higher for IV rather than lM injec-tions (Table 5).

The logistic regression on the male calf rate withPGF2a, ROUTE, and PGF2a x ROUTE in the modelyielded a non-significant interaction term (P=0.27). Re-sults of univariate logistic regression analysis are pre-sented in Tables 6 and 7. As for the fertility rates, themale calfrates are also presented for the simple effectsof each treatment group (Table 6) and for the main ef-fect of treatments, after pooling data (Table 7). ln eachtable, results are further stratified according to the in-terval (days) from PGF2a injection to breeding. Ahighermale calfrate was observed when dinoprost was injectedIV rather than lM (+26.4%, P=0.04 on 47 calves; Table6). The male calf rate when dinoprost was injected IVwas also greater than the expected population value(+33.4%, P

Table 4. Fertility rates for each treatment group (multivariate logistic regression analysis).

Dinoprost Cloprostenol

IMa IVa P IMa IVa P

Cases, n 102 99 101 100Breeding rate, % 85.9 79.3 0.23 82.8 83.4 0.90Conception rate, % 33.7 43.6 0.21 41.8 51.5 0.19Pregnancy rate, % 29.3 33.6 0.45 34.9 43.6 0.50

alM=intramuscular and IV=intravenous

Table 5. Fertility rates for the main effects oftypes ofprostaglandins (PGF2,,) and routes of injection (multivari-ate logistic regression on pooled data).

Type ofPGF2" Route of injection

Dinoprost Cloprostenol P IMa IVa P Overall

Cases, n 201 201 203 199 402Breeding rate, % 82.6 83.0 0.93 84.1 81.4 0.45 82.84Conception rate, % 38.6 46.6 0.13 38.7 46.7 0.13 42.64Pregnancy rate, % 31.4 39.2 0.14 32.5 38.1 0.23 35.32

alM=intramuscular and IV=intravenous

Table 6. Male calfrates for each treatment group and stratification according to the interval (days) from prostag-landins (PGF2,,) injection to breeding (univariate analysis).

Dinoprost Cloprostenol

IMa IVa P IMa IVa P

Abortions, n 3 2 1 1Cows sold before calving, n 3 4 2 1Missing genderb, n 3 3 1 4Set of twins, n 1 0 2 0Calves with recorded gender, n 25 22 36 34Male calf rate, % 60.0 86.4 0.04 41.7 52.9 0.34P-valuec 0.49

Table 3. Distribution, frequency, fertility rates (univariate analysis) and P-value (Chi-Square test of indepen-dence) of aH covariables.

Breeding rate Concep.t;ionrate Pregnancl ra~$(),'LLIG OA.l C- G-- 5

VARIABLE n (%) % P % P % P

Lactation number0 70(17.4) 87.1 57.4 50.01 93 (23.1) 80.7 50.7 40.92-3-4 184 (45.8) 81.5 38.7 31.5>4 55 (13.7) 85.5 0.63 23.4

Table 7. Male calfrates for the main effects oftypes ofprostaglandins (PGF2.) and routes of injection and strati-fication according to interval from PGF2(l injection to breeding (univariate analysis on pooled data).

TypeofPGF2" Routeof injection

Dinoprost Cloprostenol P IMa IVa P Overall

Calves,n 47 70 61 56 117Malecalf rate, % 72.3 47.1

l error, i.e., declaring that a difference exists when, infact, there is no such difference; 2) ~=20%, being theprobability ofmaking a type II error, i.e., declaring thatthere is no difference when a difference indeed exists;and 3) a minimum predicted difference of 5% in preg-nancy rate (i.e., 35 vs 30%)between treatments, whichis considered of clinical and economical importance.Sample size needed in each group, using those criteria,was 1,375 animaIs (Win Episcope version 2.0).

Statistical power ofa test equals 1 - ~and is equiva-lent to 80% in the ideal case with 1,375 animaIs persample. Statistical power indicates the capacity of astudy to detect a predicted difference if it really exists.With approximately 200 animaIs per sample when themain effect of treatments was considered, the statisti-cal power of the studywas 19%instead of80%. Upgrad-ing CI. to 10%is one way to increase the statistical power.Using P

(P=0.21,Table 4) for dinoprost IVcompared to dinoprostlM, yet, male calfrate was 26.4%higher (P=0.04,Table6). Thissuggests that sex-specific death offemale em-bryos after fertilization was not a contributing factor tothe male calf rate obtained with dinoprost IV:

Odds ratios in Table 8 were derived from pub-Iished data. Odds ratio is a measure of associationwhich approximates how much more likely (or unlikely)it is for the animal to be bred or pregnant when usingcloprostenol rather than dinoprost. Odds ratio higherth an 1.00 indicates better results obtained withcloprostenol compared to dinoprost and conversely ifOR is smaller than 1.00. Odds ratio allows between-studies comparisons which are independent from theaverage fertility rate obtained in each study. For ex-ample, odds ratios for breeding rate were similar, yetaverage breeding rates were quite different, being53.8% and 85.4%, respectively, in Sudweeks et apo andfor the herd C ofSalverson et a[ls (Table 8). Cloprostenolwas about 2.4 times more likely to induce an estruswhen compared to dinoprost in both studies, yet breed-ing rates for cloprostenol vs dinoprost were 65.7% vs42.7% (+23.0%) and 90.2% vs 80.0% (+10.2%), respec-tively, in Sudweeks et apo and for the herd C ofSalverson et al.ls Comparison ofresults between stud-ies is easier using an odds ratio basis (Table 8) but isnonetheless difficult due to differences in protocol,

treatmentregimen, breed, lactating status, parity ofanimaIs, etc. Great variability exists between studiesand islikely to resultfrom insufficient sample size andpresence ofconfounding factors.

It is appropriate to compare the efficacyof the twomain sources ofPGF2a, since they are commonlyinjectedalone when a functional corpus luteum is detected, es-pecially in small farms where adequate heat detec-tion is feasible. The practicing veterinarian is oftenchallenged by dairy producers about the equal efficacyof those PGF2aproducts. Dinoprost and cloprostenolhave also been used without distinction in the new pro-tocols for timing of insemination and their equal effi-cacy remains to be addressed. A larger scale study onhigh-producing dairy cattle is definitely needed to clarifythe effects of type ofPGF2a, route of injection and timeinterval from PGF2a injection to breeding on fertilityand sex ratio of calves.

Conclusion

ln conclusion, type of PGF2a used (cloprosteno1)and route ofinjection (IV)might improve conceptionrateand, hence, pregnancy rate, yet the statistical power ofthe study was too weak to detect any significant differ-ences. It can also be concluded that type of PGF2a(dinoprost), route of injection (IV) and increasing the

aD=dairy; B=beef; H=heifer; C=cow; NL=non lactatingbBR=breeding rate, CR=conception rate, PR=pregnancy ratecORhigher than 1.00 indicates better results obtained with cloprostenol compared to dinoprost and conversely if OR is smallerthan 1.00dOnly lst breeding considered; conception rate for pt breeding assumed similar to conception rate for lst and2nd breeding; sam pIesize assumed to be 72 and 73, respectively for cloprostenol and dinoprost*P

interval from PGF2a injection to breeding might skewthe sex ratio toward males.

Acknowledgements

The author wishes to thank hisformer practicingcolleagues Paul Cardyn, Robert Dodier, Pierre Brassard,Jim Porter, and Walter Canzani from the Coaticook Vet-erinary Clinic, for participating in the clinical part ofthe study, as weIl as Arnaud Renard for his input in thepreliminary statistical analysis and Claudette Cormierfor technical support. Thanks also to HerménégildeTwagiramungu and Luc Descôteaux for reviewing themanuscript and to the Upjohn company for the gener-ous donation of dinoprost used in this study.

Footnotes

a Lutalyse® - Pharmacia et Upjohn Company,Orangeville, Ontario

b Estrumate® - Coopers Agropharm, Ajax, OntarioC Beef quality assurance (BQA) guidelines now recom-mend that aIl injections be administered only infront of the point of the shoulder.

d DSA, version 4.2, ASTLQ inc., Faculté de MédecineVétérinaire, Université de Montréal, Québec, J2S7C6

e Amélioration de la Santé des Troupeaux Laitiers duQuébec

f Statistix, version 7.0, Analytical Software, Tallahas-see, FL 32317

g Win Episcope, version 2.0, developed by K. Frankenaand J.O. Goelema, from the Department ofAnimalSciences ofWageningen Agricultural University (TheNetherlands)

References

1. Bourne GR: A review of metabolism and clearance studies with14C-cloprostenol in the cow. Acta Vet Seand Suppl 77:5-9, 1981.2. Butler WR: Nutritional interactions with reproductive perfor-mance in dairy cattle. Anim Reprod Sei 60-61:449-457, 2000.3. Dobson H, Smith RF: What is stress, and how does it affect repro-duction? Anim Reprod Sei 60-61:743-752, 2000.4. Ferguson JD: Diet, production and reproduction in dairy cows.Animal Feed Sei Teeh 59:173-184, 1996.

19

5. Grohn YT, Rajala-Schultz PJ: Epidemiology of reproductive per-formance in dairy cows. Anim Reprod Sei 60-61:605-614, 2000.6. Gustafsson B, Backstrom G, Edqvist LE: Treatment of bovinepyometra with prostaglandin F2a: an evaluation of a field study.Therio 6:45-50, 1976.7. Heuwieser W,Oltenacu PA, Lednor AJ, Foote RH: Evaluation ofdifferent protocols for prostaglandin synchronization to improve re-productive performance in dairy herds with low estrus detection effi-ciency. J Dairy Sei 80:2766-2774, 1997.8. Hosmer DW,Lemeshow S:Applied Logistic Regression, ed 2, 2000.9. JohnsonAP: Response to prostaglandin therapy by lactating dairycows with unobserved estrus. Vet Med 4:555-557,1984.10. Kelton DF:Management of subestrus cows using rectal palpa-tion, a milk progesterone enzyme immunoassay and three prostag-Iandins. MSe Thesis. University of Guelph, Ontario Veterinary Gollege,1989.11. Lucy MC:Reproductive loss in high-producing dairy cattle: wherewill it end? J Dairy Sei 84:1277-1293, 2001.12. Manns JG, Hafs HD: Controlled breeding in cattle: a review. GanJAnim Sei 56:121-130, 1976.13. Maurer RR, Echternkamp SE, Wise TH: Ovarian responses tointramuscular or intravenous administration ofprostaglandin F2 al-pha in control and FSH-treated beefheifers. JAnim Sei 67:2075-80,1989.14. Pursley JR, Silcox RW,Wiltbank MC: Effect of time of artificialinsemination on pregnancy rates, calving rates, pregnancy loss, andgender ratio after synchronization ofovulation in lactating dairy cows.J Dairy Sei 81:2139-2144, 1998.15. Rorie RW:Effect of timing of artificial insemination on sex ratio.Therio 52:1273-1280, 1999.16. RoyGL, Twagiramungu H: Insemination strategies to control thesex ratio in cattle. A.E. T.E. 16'"Seientifie Meeting, Santander, Spain,2000.17. Ryan DP, Boland MP: Frequency ofbirths among Holstein-Frie-sian cows in a warm dry climate. Therio 36:1-10,1991.18. Salverson RR, DeJarnette JM, Marshall CE, Wallace RA: Syn-chronization of estrus in virgin beef heifers using melengestrol ac-etate and PGF2alpha: an efficacy comparison of cloprostenol anddinoprost tromethamine. Therio 57:853-858, 2002.19. Seguin B, Momont H, Baumann L : Cloprostenol and dinoprosttromethamine in experimental and field trials treating unobservedestrus in dairy cows. Bov Praet 20:85-90, 1985.20. Sudweeks EM, Randel RD, Tomaszewski MA:Estrous synchroni-zation after prostaglandin injection in dairy heifers. J Dairy Sei 66(suppl1):231 (abstract), 1983.21. Turner RB, Peterson GA, Davis ME, Wilson GR, Irvin KM, ForryJTT : Synchronization of estrus in beef cows and heifers withfenprostalene, cloprostenol sodium and prostaglandin F2 alpha. Therio28:15-24,1987.22. Wehner GR, WoodC, Tague A, Barker D, Hubert H: Efficiency ofthe OVATECunitfor estrus detection and calf sex control in beef cows.Anim Reprod Sei 46:27-34,1997.23. Wenkoff MS: The management of drug-induced manipulation ofthe estrous cycle in normal cows and heifers. Gan Vet J 28:366-373,1987.

THE BOVINE PRACTITIONER-VOL. 37, NO. 1