HAL Id: hal-00902439https://hal.archives-ouvertes.fr/hal-00902439

Submitted on 1 Jan 1996

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

The use of an experimental metritis model to studyantibiotic distribution in genital tract secretions in the

eweCc Cester, Jc Béguin, Pl Toutain

To cite this version:Cc Cester, Jc Béguin, Pl Toutain. The use of an experimental metritis model to study antibioticdistribution in genital tract secretions in the ewe. Veterinary Research, BioMed Central, 1996, 27(4-5), pp.479-489. �hal-00902439�

Original article

The use of an experimental metritis modelto study antibiotic distribution in genital tract

secretions in the ewe

CC Cester JC Béguin 2 PL Toutain

1 Unité associée Inra de physiopathologie et toxicologie expérimentales,École nationale vétérinaire, 23, chemin des Capelles;

2 Laboratoire de microbiologie, Rhône-Mérieux, 4, chemin du Calquet, 31076 Toulouse, France

(Received 15 November 1995; accepted 26 March 1996)

Summary &horbar; The influence of experimentally-induced metritis on spiramycin disposition in genitalsecretions was investigated in six ovariectomized ewes. A crossover study design was selected tocompare control with metritis pharmacokinetics. A clinically-relevant metritis was obtained under pro-gestagen priming by inoculation in the uterine lumen of a bacterial suspension of Actinomyces pyogenesand Fusobacterium necrophorum Ewes were given a single iv administration of spiramycin at a doseof 20 mg·kg-!. Plasma and genital secretions were regularly sampled up to 96 h post-injection and spi-ramycin activity was measured using a microbiological method. Experimental metritis did not affectplasma spiramycin disposition and the antibiotic was more concentrated and lasted longer in genitalsecretions than in plasma regardless of the animal’s state of health. The area under the concentra-tion-time curve of spiramycin in genital secretions was twofold higher (p < 0.05) in infected than in healthyewes (3361 ± 112 JJgohog-1 and 175 ± 41 JJgohog-1 respectively). The mean residence time of spi-ramycin in genital secretions was significantly longer in diseased ewes (32 ± 4 h) than in control ewes(23 ± 4 h). The maximum concentration of spiramycin in genital secretions was equal for both studiesbut occurred later in infected ewes (2.7 ± 1.0 h versus 8.6 ± 4.5 h). It was concluded that a uterine infec-tion had a marked influence on the disposition of spiramycin in genital tract secretions and that this uter-ine infection model in the ewe merits consideration for the study of drug treatments of genital tractinfection.

experimental metritis / genital secretion / pharmacokinetic / spiramycin

*

Correspondence and reprints

Résumé &horbar; Utilisation d’un modèle de métrite expérimentale pour l’étude de la distribution desantibiotiques dans les sécrétions génitales chez la brebis. L’influence d’une métrite expérimentalesur le passage de la spiramycine dans les sécrétions génitales a été étudiée chez six brebis ovariec-tomisées. Un plan expérimental croisé a permis de comparer les cinétiques réalisées chez les animauxtémoins à celles réalisées chez les mêmes animaux après induction d’une métrite. La métrite expéri-mentale a été obtenue chez des brebis sous imprégnation progestéronique en introduisant dans lalumière utérine une suspension bactérienne dActinomyces pyogenes et de Fusobacterium necro-phorum. Les brebis ont reçu une administration intraveineuse de spiramycine (20 mg kfT 1) ; le plasmaet les sécrétions génitales ont été régulièrement recueillis pendant les 96 heures qui ont suivi l’admi-nistration. L’activité de la spiramycine dans les échantillons a été mesurée selon une méthode micro-biologique. La métrite expérimentale n’a pas modifié la cinétique plasmatique de la spiramycine ; lesconcentrations d’antibiotique ont été élevées et se sont maintenues plus longtemps dans les sécrétionsgénitales que dans le plasma, chez la brebis saine et infectée. Les aires sous la courbe des concen-trations de spiramycine dans les sécrétions génitales ont été deux fois plus importantes chez les bre-bis infectées (361 ± 112 pg.h-_q-1) que chez les témoins (175 ± 4? pg.h-_q-1). Le temps moyen derésidence de la spiramycine dans les sécrétions génitales a été plus long chez les animaux infectés (32f 4 heures) que chez les sains (23 ± 4 heures). Les concentrations maximales de spiramycine dansles sécrétions génitales ont été similaires mais sont apparues plus tard chez les brebis infectées (2,7 7f 1,0 versus 8,6 ± 4,5 heures). Cette étude montre que la métrite expérimentale modifie le passage dela spiramycine dans les sécrétions génitales et que ce modèle d’infection utérine mérite d’être retenupour l’étude des médicaments utilisés dans le traitement des métrites.

métrite expérimentale / sécrétion génitale l pharmacocinétique / spiramycine

INTRODUCTION

To obtain effective treatment of a genitaltract infection, an antibiotic should be admin-istered at the most appropriate dose, dosinginterval and route of administration (localversus general) to provide drug concentra-tions at the uterine site of infection whichmaximize bacterial killing. The intrauterineroute is usually selected in veterinarymedecine, because it is generally consid-ered to be the most convenient and because

it is assumed that the antibiotic concentrationwill then be higher at the infection site. Infact, when bacteria spread out of theendometrial lumen and invade the

myometrium and oviducts, a general route ofadministration is in order (Gustafsson, 1984;Bretzlaff, 1986; Gilbert and Schwark, 1992).The selection of an appropriate dosage fora systemic route of absorption requireskinetic information on antibiotic disposition inplasma and its passage through the uter-ine barrier. Most experiments published todate have studied uterine antibiotic dispo-

sition in healthy animals although it is recog-nised that metritis can modify the ability of adrug to cross the uterine barrier. Genitalinfections may increase blood flow to theinflamed uterus or damage the endometrium(Ayliffe and Noakes, 1982). The purpose ofthe present experiment was to develop anexperimental animal model of metritis todocument the passage of an antibiotic

through the uterine barrier when a systemicroute of administration was selected.

MATERIALS AND METHODS

Animals

Six cyclic, multiparous Romanov-cross ewes(weighing 43 ± 6 kg) were used. At least onemonth before the experiments, the ewes wereovariectomized under aseptic surgical conditionsand a silicone catheter (Rhodorsil; od, 3 mm) wasinserted into the upper third portion of each uter-ine horn via a small puncture hole in the uterinewall. The catheters were maintained inside the

uterus with ligatures and the free ends of thecatheters were exteriorized to the flank of the ani-mals and clamped. After surgery, the animalswere placed in individual cages and were pro-vided oats, hay and water ad libitum.

The body temperature and the pH of the gen-ital secretions were measured with an electronicthermometer and electronic pH-meter respec-tively, in three control and three metritis ewesonce a day before spiramycin administration andfor the four days following antibiotic administration.

Experimental design

Ewes were randomly assigned to groups A (n =3) or B (n = 3) in a crossover study with respectto uterine pathologic status. In group A the con-trol disposition of spiramycin was investigatedbefore the disposition during metritis, and viceversa in group B. A washout period of five weekswas observed between the two studies. After

experiments performed during metritis, eweswere given a daily intramuscular administrationof penicillin-dihydrostreptomycin combination(Mixtencilline, Rhône-Mérieux, Toulouse, France)at a dose of 106 international units (IU) in totofor five days.

Control pharmacokinetics

Control kinetics of spiramycin were examined inuninfected ewes under oestrogen priming. Theewes were treated with progestagen by meansof a 40-mg fluorogestone acetate (FGA) sponge(Chronogest, Intervet, Angers, France) placed inthe vagina for 14 days. Twenty four hours aftersponge withdrawal, 30 pg of 17 P oestradiol(oestradiol Benzoate, Intervet, Angers, France)was injected intramuscularly. The administrationof 17 j3 oestradiol benzoate was continued dailyfor five days.

Induction of metritis

The procedure used for metritis induction wasadapted from a method described by Farin et al(1989) who were studying the role of Actinomycespyogenes and Gram-negative anaerobic bacteria

in producing a pyometra. The day of the first inoc-ulation with the bacterial suspension was desig-nated as Day 0. On Day-4, a polyurethane spongeimpregnated with 40 mg fluorogestone acetatewas inserted into the vagina. One day before thefirst inoculation (Day-1), the uterus was infusedwith 5 mL per uterine horn of an irritating iodinesolution via the uterine catheters. Both ovine uter-ine horns were inoculated on three consecutive

days (Day 0, Day 1 and Day 2) with 5 mL perhorn of bacterial suspension (Actinomyces pyo-genes: 5 x 10! colony-forming units-mL-1(cfuomL -1), ), Fusobacterium necrophorum: 5 x 10!cfuOmL-1). ).

A pyogenes used in the inocula was obtained

using swabs from the uterus of a cow with metri-tis and F necrophorum was isolated from a sheepwith foot rot. The bacteria were stored at -80 °Cuntil used to prepare the bacterial suspensionwhich was then infused into the uterus. After thaw-

ing, 0.1 mL of each bacterial strain was streakedon to Columbia Blood Agar (Bio-Mérieux, Marcy-I’!toile, France) and incubated under anaerobicconditions for 72 h at 37 °C. After purity controland identification by standard methods, the bac-teria were transferred into a trypcase-soy broth(Bio-Mérieux, Marcy-l’!toile, France) and dilutedto a final concentration of 108 cfu·mL-!. The final

suspension was prepared with an equal volumeof each bacterial suspension, homogenized anddispensed into 30-mL sterile flasks. These sus-pensions were stored at -80 °C until they wereinfused into the ovine uterus. Before inoculation,the bacterial suspension was thawed for 3 h at37 °C.

Drug administrationand samples collection

One day before the experiments, a catheter filledwith heparinized 0.9% saline solution was insertedinto each jugular vein. Spiramycin (Suanovil 20,Rhone-Merieux, Toulouse, France) was injectedintravenously as a single dose of 20 mg-kg-1 via

one of the catheters. The control kinetic studybegan 15-18 h after the first administration ofbenzoate oestradiol while the metritis kinetic studystarted 48 h after the last bacterial uterine infusionand 3 h after withdrawal of the FGA vaginalsponge.

Blood samples were collected from the othercatheter into heparinized tubes before dosing

(baseline sample, time 0) and then 1, 2, 4, 8, 15,30, 45, 60 and 90 min and 2, 4, 6, 8, 9, 10, 24, 30,48, 54, 72 and 96 h after antibiotic injection.Plasma was separated from blood by centrifu-gation and frozen until assayed.

Genital tract secretions were withdrawn usinga method adapted from that described by Lindsayand Francis (1968). In brief, polyurethanesponges (Chrono-Gest placebo, Intervet, Angers,France) were inserted into the vagina of the ewesfor one hour. Genital secretions samples wereobtained before injection and 1, 3, 6, 9, 24, 30, 48,54, 72 and 96 h after antibiotic injection. Thesecretions were squeezed out of the sponge andstored at -20 °C until being assayed for spi-ramycin.

Analytical assay

The antibiotic activity of spiramycin in plasmaand genital secretions was measured by agar geldiffusion using Sarcina lutea ATCC 9341 as thetest organism according to the proceduredescribed by the European Pharmacopoeia(1980). Just before the analytical assay, the gen-ital secretions samples were centrifuged at 6000 gfor 10 min and the supernatant removed. Spi-ramycin concentrations were measured in thesupernatant obtained after centrifugation of gen-ital tract secretions. Plasma and genital secre-tions were assayed simultaneously with knownspiramycin standard solutions prepared in plasmaand genital secretions respectively, obtained fromuntreated ewes. The level of quantitation was0.0625 mg·L-! for both plasma and uterine secre-tions and the coefficient of variation of repeatabilityand reproducibility was less than 18%.

Pharmacokinetic analysis

The plasma concentrations of spiramycin wereanalysed with a program for non-linear regres-sion analysis adapted from Multi (Yamaoka et al,1981). Each data point was weighted with thereciprocal of the fitted value:

where lN; was the weight, and !i the fitted valueof the Rh datum. Initial estimates of the parame-ters of the equation describing the data were

obtained by linear regression methods (Gibaldiand Perrier, 1982). The following exponentialequation was fitted to the data:

where Ct (mgoL-1) represented spiramycin activ-ity at time t, Yi (mgoL-1) were the coefficients and,li (h-1) the exponents. The selection of the num-ber of exponents of equation [2] was determinedaccording to the Akaike information criterion(Yamaoka et al, 1978). The pharmacokineticparameters (elimination half-time, body clear-ance, volume of distribution) were calculatedusing the classic equations associated with com-partmental analysis (Gibaldi and Perrier, 1982).Plasma clearance (C! was calculated using theequation:

where D was the dose, and AUCO-inf, the areaunder the plasma concentration-time curve cal-culated using the trapezoidal rule from time 0 toinfinity. The extrapolated part of the curve wascalculated between the last experimental pointand infinity using the equation:

where Ciast was the concentration of the last sam-

ple and llz, the slope of the terminal phase. Themean residence time (MRT) was calculated bythe linear trapezoidal rule with and without extrap-olation to infinity.

Equations describing the fate of drugs admin-istered by a non-vascular route, ie, including aninvasion phase and a lag-time (fia9) before theonset of invasion, were fitted to the spiramycinconcentrations in genital secretions:

In equation [5], Ct (¡.¡gog-1) represents spi-ramycin activity in genital secretions at time t, Yiand Y2 (¡.¡gog-1) are coefficients; 7!! and !2 (h-!)are exponents; 7!&dquo; (h-!) is the invasion rate con-stant. No weighting factor was used to fit theequation to genital secretions data.

The spiramycin concentrations were also anal-ysed using a non-compartmental approach. The

AUC and MRT for plasma and genital secretionswere selected to evaluate the extent of the diffu-sion from blood to genital secretions and to qual-ify the influence of metritis on spiramycin dispo-sition in genital tract secretions. The AUCo_!astand MRTO-Clast were calculated using the trape-zoidal rule from time 0 to the last concentrationmeasured in plasma or in secretions. The secre-tions-to-plasma concentration ratio was calcu-lated using the AUCO-Clast for both plasma andgenital secretions for both kinetic studies.

Statistical analysis

The pharmacokinetic parameters derived fromall plasma and genital tract secretion data wereexpressed as mean ± standard deviation (SD)for the six ewes. Harmonic means were calcu-lated for the half-times of invasion and elimination;standard deviations were computed using a jack-knife technique (Lam et al, 1985).

Statistical analysis was performed using Stat-graphics (Statistical Graphics Corporation, 1988).The pharmacokinetic parameters of the control

and metritis kinetic studies were compared usinga four-factors ANOVA (pathological status, period,sequence and animals nested in sequence). Thelevel of significance was 0.05. The plasma andgenital secretions-derived parameters were com-pared using a two-factor ANOVA (animal, bio-logical matrix).

RESULTS

Animals

Clinically-relevant metritis was confirmed inall ewes by the presence of purulent vaginaldischarge while control ewes showed typicalclear oestral mucus. Body temperature wasnot modified in metritis ewes and behaviour

looked normal throughout the experiment.The pH values of genital tract secretionswere 8.4 ± 0.1 and 8.3 ± 0.1 in control and

metritis ewes respectively.

Plasma kinetics

Figure 1 shows the mean semi-logarithmicplots of spiramycin activity in plasma andgenital secretions in six ewes with and with-out uterine infection.

According to the Akaike information cri-terion, a triexponential equation best fittedthe plasma experimental data for each ewe(equation [2]). Consequently, a three-com-partmental model was used to analyse theplasma disposition of spiramycin. Meanpharmacokinetic values for the six ewes aregiven in table I. Comparison of plasma phar-macokinetic parameters did not show anysignificant difference between the twogroups (p > 0.05, ANOVA). The half-times ofelimination were (t1/n) 17.0 ± 3.3 and 18.0

± 3.9 h and the mean residence times

(MRT) were 10.5 ± 3.8 and 12.7 ± 3.2 h forcontrol and metritis kinetics respectively.These parameters indicated a slow disap-pearance of spiramycin from plasma. Thesteady state volume of distribution ( VSS) washigh (6.0 ± 1.7 and 7.9 ± 1.9 L·kg-!), sug-gesting considerable distribution of spi-ramycin in the tissues.

Genital tract secretion parameters

The area under the time-concentration

curves and the mean residence time were

significantly higher in genital secretions thanin plasma (p < 0.001, ANOVA) whateverthe health condition of the ewes (tables 11

and 111). The area under the time-concen-tration curves calculated from genital secre-tion data was twofold higher for infectedewes (p < 0.05, ANOVA): ie, 175 ± 41 and361 ± 112 2 pg.h.g-I respectively (table II).Similarly, the mean residence time of spi-ramycin in genital secretions was signifi-cantly longer in ewes with metritis than incontrol ewes (32.5 ± 3.8 versus 22.8 ± 4.3 h,p < 0.001 ANOVA) (table III).

The compartmental pharmacokineticanalysis showed that a biexponential equa-tion best fitted the genital secretions datawith an invasion phase and with a lag-timefor some data sets (n = 2 for healthy ewesand n = 4 for metritis ewes) (equation [5]).The mean values for genital secretionparameters are given in table IV. The pas-sage of spiramycin from blood to genitaltract secretions (half-time of invasion, t1/2ÀA)was markedly slower in ewes with metritisthan in control conditions (p < 0.05,ANOVA). This was confirmed by a maximalconcentration (equal for both groups) that

occurred later in metritis ewes (8.6 ± 4.5 h)than in healthy ewes (2.7 ± 1.0 h). The dif-ference between these values was signifi-cant for p = 0.053 (ANOVA). The eliminationrate constants for uterus (!) were signifi-cantly lower in infected than in control ewes(p < 0.001, ANOVA), leading to a twofoldlonger half-time of elimination (t1/2À2) of spi-ramycin from the uterus (28.2 ± 9.2 versus50.2 ± 10.7 h in control and metritis ewes

respectively).

DISCUSSION

In the present experiment, genital tractsecretions were chosen to assess diffusion

from blood to the uterine lumen of an antibi-

otic administered by a systemic route.Female genital pathogens invade the deeplayers of the uterine wall from the uterinelumen and the presence of antibiotic in uter-

ine and cervical secretions after an iv admin-

istration guarantees the passage of drug

through the uterine barrier and its presencethroughout the biophase.

The procedure of intrauterine infusion ofbacteria described in this experiment sucess-fully induced uterine infection in all ewes.The choice of A pyogenes and F necropho-rum in our experimental model was guidedby the fact that these bacteria have oftenbeen isolated from the uterus of diseasedcows and ewes (Bane, 1980; Roberts,1967b; Olson et al, 1984) and A pyogenes isconsidered as the most harmful pathogenin metritis. Moreover, several investigationshave demonstrated a synergistic relation-ship between A pyogenes and F necropho-rum in producing pathologic changes (Sokkaret al, 1980; Roberts, 1967a; Farin et al,1989). Progestagen priming and iodine treat-ment were used respectively to reducedefense mechanisms and damage theendometrium, thereby promoting bacterialinvasion of the genital tract. The presenceof metritis was ascertained by vaginal specu-lum examination: infected ewes had cloudy,malodorous genital secretions with small

clumps of pus whereas the oestral mucusof healthy ewes was clear. Infected ewesshowed no modification in behaviour and

general health status and rectal tempera-ture remained within the normal range.

The iv administration of spiramycin pro-duced higher and longer concentrations ingenital secretions than in plasma regardlessof animal health status. These results are in

agreement with those reported for spiramycinin previous works (Cester et al, 1990, 1992).The area under the time-concentration curvefor genital secretions and the secretion-to-plasma concentration ratio were higher inmetritis than in control ewes. Similarly, exper-imental metritis increased spiramycin per-sistence in genital secretions. The statisti-cal analysis did not reveal any significantdifference in the plasma pharmacokineticparameters between control and metritisewes. The increased amounts and longerpersistence of spiramycin in genital secre-tions of metritis ewes cannot be attributedto modification of the spiramycin plasma dis-position. This is consistent with the fact that

experimental metritis did not induce feversyndrome or impair general health status. It

may be assumed, therefore, that the modi-fication of spiramycin disposition in the gen-ital secretions was due solely to a new localcondition induced by the metritis. This can-not be attributed to a modification in pH par-titioning with respect to genital tract infec-tions. The pH values of genital secretionswere in fact similar (8.4 ± 0.1 versus 8.3 ±0.1) in both diseased and healthy oestrogenpriming ewes and cannot be advocated as afactor influencing the passage of antibioticfrom blood to genital secretions.

The compartmental analysis suggestedthat the experimental metritis was associ-ated with a decrease in spiramycin transferthrough the uterine barrier. The passage ofspiramycin from blood to genital secretionswas slower in diseased ewes as indicated bya longer invasion half-time (t1m.A) and alater time of occurrence of the maximal con-

centration (Tmax) of spiramycin in the secre-tions. The faster passage of drug from bloodto the uterine lumen in control ewes was

certainly due to an increased genital bloodflow. It is currently recognized that oestrogenincreases the passage of antibiotics throughthe uterine barrier by increasing uterineblood supply or endometrial capillary (Bret-zlaff, 1986). Our control animals were underoestrogen priming because the direct col-lection of genital secretions is impossible inthe ewe under progestagen. In a previouswork (Cester et al, 1992), the passage ofspiramycin into uterine secretions was inves-tigated in uninfected oestrogens and pro-gestagen ewes: the amounts of spiramycinthat penetrated into uterine secretions werehigher under oestrogen priming than underprogestagen priming. In the present exper-iment, the amounts of spiramycin in genitalsecretions were higher in infected pro-gestagen ewes than in uninfected oestro-

gen ewes, this steroid environment beingconsidered as the most favourable steroidconditions of a good uterine penetration ofspiramycin. Then the increased spiramycinamounts in genital secretions may beattributable to infection and not to steroidhormones. Besides, it is known that oestro-

gens increase the uterine defense mecha-nisms and experimental metritis should notbe induced under oestrogen priming. In aprevious work we have shown that spi-ramycin uterine availability, but not its resi-dence time in genital secretions, wasincreased in ewes under oestrogen primingas compared in ewes under progestagenpriming (Cester et al, 1992) and it was con-cluded that this increase was certainly dueto an increased blood flow and/or uterinebarrier permeability attributable to oestro-gens. In this experiment, it can therefore be

assumed that the uterine blood flow was

probably increased less by inflammation fol-lowing bacteria inoculation in our metritisconditions than by the oestrogen priming inour controls. The greater amounts of spi-ramycin in genital tract secretions of metri-tis ewes cannot be explained, therefore, bya greater passage of spiramycin from theblood to the uterus but more likely by lessspiramycin passing from the uterus to theblood. This is consistent with the half-time ofelimination (t1/2I-2) and MRT for genitalsecretions which are longer in diseasedewes. This might be attributable to anincreased binding of spiramycin to uterinecontents (eg, secretions, pus, subendome-trial tissues) or to the penetration of spi-ramycin into the polynuclear neutrophil pop-ulation which dramatically increases duringinfections. It should be noted that consid-

erable leucocyte passage into the uterinelumen occurs in the presence of A pyo-genes (Constantin, 1987), and that spi-ramycin penetrates easily into macrophagesand polynuclear cells as has been demon-strated for the respiratory tract (Zeneberghand Trouet, 1982; Pocidalo et al, 1985; Harfet al, 1988).

The efficacy of a treatment using a bac-teriostatic antibiotic requires the persistenceof sufficient concentrations in the genitaltract for a sufficient time to allow the elimi-

nation of pathogens by uterine defencemechanisms. The minimal inhibitory con-centration (MIC) of spiramycin for A pyo-genes, considered to be the most harmful

pathogen in metritis, is approximately3 mg*L-1 (Atkinson, 1986). In the presentstudy, spiramycin concentrations in genitaltract secretions remained close to 3 Ng·g-1for approximately 48 h. These results sup-port the use of the intravenous route for spi-ramycin administration and it can be

assumed that the systemic route is likely toprovide a good distribution of the drug ingenital tract tissues and fluids, even of dis-eased animals.

The present experiment has clearlydemonstrated that an experimentally-induced metritis can easily be obtained inthe ewe and that the uterine disposition of asystemic administration of antibiotic waslargely influenced by genital tract infection.It can be concluded, therefore, that a rele-vant antibacterial treatment for metritis can-not be based only on kinetic results involv-ing healthy cycling animals. It is likely thatthe same considerations hold for genitaltract infections of women.

ACKNOWLEDGMENTS

Special thanks to N Gautier for the help duringthis study.

REFERENCES

Atkinson BA (1986) Species incidence and trends ofsusceptibility to antibiotics in the United States andothers countries: MIC and MBC. In: Antibiotics in

Laboratory Medecine (V Lorian, ed), Williams &

Wilkins, Baltimore, 1110 0

Ayliffe TR, Noakes DE (1982) Effects of exogenousoestrogen and experimentally induced endometritis

on absorption of sodium benzylpenicillin from thecow’s uterus. Vet Rec 110, 96-98

Bane A (1980) Microbiology of the genital tract: etiol-ogy of genital infections. Proc IXth Int Cong AnimReprod and AI, Editorial Garsi, Madrid, 473-484

Bretzlaff KN (1986) Factors of importance for the dis-position of antibiotics in the female genital tract. In:Current Therapy in Theriogenology (DA Morrow,ed), 2nd ed, Saunders Co, Philadelphia, 43-53

Cester C, Laurentie M, Garcia-Villar R, Toutain PL (1990)Spiramycin concentrations in plasma and genitaltract secretions after intravenous administration inthe ewe. J Vet Pharmacol Ther 13, 7-14 4

Cester C, Dubech N, Toutain PL (1992 ) Effect of sexualsteroid hormones on spiramycin disposition in gen-ital tract of the ewe. J Pharm Sci 81, 33-36

Constantin A. (1977) Pathogenese et traitement desendometrites et des m6trites chez la vache. Bull

Group Tech Vet1-18 8

European Pharmacopoeia (1980) Conseil de I’Europe :Maison-neuve SA. 2nd ed; Vil. 1.3 and Vlll.4.-6

Farin PW, Ball L, Olson JD, Mortimer RG, Jones RL,Adney W, McChesney AE.(1989) Effect of Actino-myces pyogenes and Gram-negative anaerobic bac-teria on the development of bovine pyometra. The-riogenology 31, 979-989

Gibaldi M, Perrier D (1982) One compartment model.In: Pharmacokinetics (J Swarbrick, ed), 2nd ed, MDekker, New York, 1-43

Gilbert RO, Schwark WS (1992) Pharmacologic con-siderations in the management of peripartum con-ditions in the cow. Vet Clin North Am Food Anim

Pract 8, 29-56

Gustafsson BK (1984) Therapeutic strategies involvingantimicrobial treatment of the uterus in large ani-mals. J Am Vet Med Assoc 185, 1194-1198

Hart R, Panteix G, Desnottes JF, Diallo N, Leckercq M(1988) Spiramycin uptake by alveolar macrophages.J Antimicrob Chemother 22 (Suppl B), 135-140

Lam FC, Hung CT, Perrier DG (1985) Estimation of vari-ance for harmonic mean half-time. J Pharm Sci74,229-231

Lindsay DR, Francis CR (1968) Cervical mucus mea-surement in ovariectomized ewes as a bioassay ofsynthetic and phyto-oestrogens. Aust J Agric Res19, 1069-1076

Olson JD, Ball L, Mortimer RG, Farin PW, Adney WA,Huffman EM (1984) Aspects of bacteriology andendocrinology of cows with pyometra and retainedfetal membranes. Am J Vet Res 5, 2251-2255

Pocidalo JJ, Albert F, Desnottes JF, Kernbaum S (1985)Intraphagocytic penetration of macrolides: in vivocomparison of erythromycin and spiramycin. JAntimi-crob Chemother 16 (suppl A), 167-173

Roberts DS (1967a) The pathogenic synergy ofFusiformis necrophorus and Corynebacterium pyo-genes. I. Influence of the leucocidal exotoxin of F

necrophorus. Br J Exp Pathol48, 665-673

Roberts DS (1967b) The pathogenic synergy ofFusiformis necrophorus and Corynebacterium pyo-genes. I The response of F necrophorus to a filter-able product of C pyogenes. Br J Exp Pathol 48,665-673

Sokkar SM, Kubba MA, Al-Augaidy F (1980) Studies onnatural and experimental endometritis in ewes. VetPatholl7, 693-698

Statgraphics Version 5 (1991) Statistical Graphics Cor-poration, STSC Inc, Rockville, MD, USA.

Yamaoka K, Nakagawa T, Uno T (1978) Application ofAkaike’s information criterion (AIC) in the evaluationof linear pharmacokinetic equations. J PharmacokinetBiopharm 6, 165-175

Yamaoka K, Tanigawara T, Nakagawa T, Uno T (1981)A pharmacokinetic analysis program (MULTI) formicrocomputer. J Pharmacobiodyn 4, 879-885

Zenebergh A, Trouet A (1982) Cellular pharmacokinet-ics of spiramycin in cultured macrophages. AnnImmunol (Institut Pasteur) 133 D, 235-244