A PATHOLOGICAL AND BACTERIOLOGICAL ABATTOIR SURVEY OF THE REPRODUCTIVE TRACTS OF MERINO EWES IN WESTERN AUSTRALIA N. R. ADAMS,B.V.Sc., Ph.D.
CSIRO, Division of Animal Health, Private Bag, P.O. Wembley, Western Australia, 6014 Mraiuction
Lamb marking rates in Western Australia are consistently below the national average, due primarily to low twinning rates, and the failure of a high proportion of ewes to bear a lamb (Chopping and Lindsay 1970). Various reasons have been proposed for this, including the nutritional conditions during mating, microbial infections of the ewe, and clover disease infertility. However, data are lacking on the importance of any of the suggested causative factors. Statistical evidence suggests that lamb marking rates are decreased in the subterranean clover areas (Lightfoot 1972), although clinical cases of clover disease with very low lambing rates and dystocia are now extremely rare throughout most of Western Australia. Although the flora of the cervix has been studied in the ewe (Zaki and Saber 1962; Quinlivan 1970), the uterine flora, and its effect on fertility in the ewe, has not been examined. This paper records the results of a pathological and microbiological abattoir survey of genitalia of ewes from commercial flocks. Materials and Methods Samples were collected between March and May in 1972 and 1973, to avoid the effects of intake of green clover. Clover pastures in Western Australia generally lose their oestrogenicity by December, when the clover dries off (Rossiter, 1970). Collections were made from groups of breeding ewes during visits on random days to the Perth abattoirs, and generally consisted of ewes culled for age. There was no selection of ewes sampled on the basis of previous fertility history. Ewes which were obviously pregnant were not examined. Where possible, the property of origin was traced, and the manager interviewed. The ewes came from properties widely dispersed throughout the 380-760 mm rainfall belt in south Western Australia, and grazed subterranean clover pastures generally typical of the district. Between 7 and 14 ewes were sampled from 18 different sources, so that a total of 194 tracts was examined macroscopically and histologically.. Of these, 145 from 15 of the sources were also e x m e d bacteriologically. An additional 293 genital tracts were collected from 4 of the 18 sources. They came from flocks of unknown reproductive history, and the additional numbers resulted from enthusiasm on the part of the abattoir workers. These tracts were examined macroscopically, an& 72 of them were also examined bacteriologically. Thus a total of 487 tracts (194 293) were examined macroscopically, and 217 (145 72) were examined bacteriologically.
+ +
Australion Veterinary Journal, Vol. 51. July, 1975
At the abattoirs, reproductive tracts were wrapped individually in paper towels and chilled on ice. Within two hours of death both uterine horns were opened aseptically in the posterior third, and a sterile bacteriological swab passed through the lumina in order to collect material for bacteriological examination. The numbers of cysts in the uterus, cervix and vagina were counted, and classified as marked (> 10 cysts), moderate (5-10 cysts), slight (< 5 cysts) or absent. Half of each ovary, portions of isthmus and ampulla from each oviduct, the mid part of each uterine hoTn, the mid part of the cervix, and the anterior vagina from each ewe were fixed in Bouin’s solution, and sections were. stained with haematoxylin and eosin. In some specimens, the vagina was not removed at the abattoirs, and could not be examined. Material from the bacteriological swabs was inoculated onto sheep blood agar plates, and the plates were incubated in a candle jar at 27°C. In addition, uterine contents from 100 of the 217 ewes were inoculated onto sheep blood agar plates which were then incubated at 37°C in an anaerobic atmosphere containing 5 % carbon dioxide. After incubation, the number of colonies was counted, and isolates were identified by the methods described by Cowan and Steel (1965). Usually the organisms were classified only by genus. Pathological and bacteriological analyses were carried out separately and the results collated at the end of the study, and analysed statistically by the Chi-square test. Results
Macroscopic Observations Massive adhesions involving the entire internal genitalia were observed in 8 ewes (1.6%). Adhesions of the ovarian bursa were seen in a further 5 ewes ( l % ) , and hydrosalpinx in 4 of these. Ovaries were judged as inactive (no corpus luteum or large follicle) in 21 ewes (4.3% 1. Most ewes with inactive ovaries came from groups with very poor body condition. Cystic endometrium was observed in ewes from 11 of the 18 groups. A total of 88 ewes ( 1 8 % ) had cystic endometrium, which was classified as marked in 30, moderate in 29, and slight in 29, Macroscopic cysts were observed in the cervix of ewes from 13 groups, with a total of 95 ewes (19.5%) af3ected. A total of 159 ewes (32.7%) had macroscopic cysts in either uterus or cervix. Pyometron was seen in 4 ewes, and hydrometron in 3 ewes. Endornetritis was observed in 3 other ewes, and in 1 of these was accompanied by embryonic resorption. Embryonic 35 1
remnants were observed in 2 other ewes. Purulent vaginitis was observed in 8 ewes, generally associated with grass seeds in the anterior vagina or posterior cervix. Unilateral aplasia of the uterine horn and caudal part of the oviduct was observed in 2 ewes, 1 on the right side, and the other on the left. One of there ewes was pregnant. One hundred and twenty-three ewes were examined for macroscopic remnants of the Wolffian duct. Parovarian cysts were found in 25 ewes (20.3%) including 1 1 on the right side, 10 on the left, and 4 bilateral. Dilated Gamer’s ducts were visible in the vagina of 16 ewes 1 3 % ) .
Microscopic Observaticns Ovary - Most ovaries appeared normal histologically. Fibrous adhesions were seen on the ovaries of those ewes in which bursa1 adhesions had been observed macroscopically. The hilar region of 5 ovaries (1.3% ) contained small cysts lined with ciliated columnar epithelium. Oviciuct - Intraepithelial cysts, often in large numbers, occurred in the ampulla of 116 ewes (61 9% ). There was no statistically significant association between oviductal intraepithelial cysts and salpingitis, endometritis or cervical cysts. Hydrosalpinx was observed in 6 ewes and was biIatera1 in 2 of these. Mild salpingitis, consisting of sniall accumulations of neutrophils, or of lymphocytes and plasma cells with epithelial degeneration, occurred in 14 ewes ( 7 % ) . In 3 ewes, there was a heavy infiltration of the oviductal tissue by eosinophils. Ulerus - Microscopic cysts were seen in the uterus of 28 ewes (14%), and dilated glands which appeared to be precursors of cysts were found in 68 ewes (35% 1. The dilated glands occurred mainly in small localized areas containing many gland buds, deep in the mucosa. Fibrosis wiis observed around blood vessels and in the muscle layers in some ewes, but periglandular fibrosis did not appear to be involved in the pathogenesis of gland dilation. Uterine glands were abundant in some ewes, and adenomyosis was observed in the uterus or cervix of 28 ewes ( I4V ) . Squamous metaplasia of the uterine epithelium was observed in one ewe. Acute endometritis was diagnosed in 29 ewes ( 15 L/r 1. in 10 of which it was pyogenic. Eschericliia coli was recovered from all ewes with pyogenic endometritis. and was frequently associated with streptococcus or corynebacteria. Chronic endometritis indicated by the presence of lymphoid nodules or large numbers of plasma cells in thc lamina propia, was observed in 37 ewes (19O’r ) , but bacteria could only be recovered 352
from 53% of these and were predominantly corynebacteria. Chronic metritis was observed in 3 ewes, and E . coli was recovered from all of these. There were associations in ewes between the presence of endometritis and uterine cysts or dilated glands (Xz1 = 13.8, P < 0.01), and between the presence of endometritis and microscopic cysts in the cervix (Xyl = 17.25, P < 0.00 1 ) . Cervix - Microscopic cysts were noted in the cervix of 70 ewes (37% ), and frequently were accompanied by an increase in glandular tissue in the cervix. Squamous metaplasia of the cervical epithelium occurred in 70 ewes, chiefly in the crypts of the posterior cervix. Hyperplastic goblet cells were seen in the cervix of 64 ewes ( 3 3 % ) , frequently associated with squamous metaplasia. In these ewes, goblet cells packed with mucus occurred in large numbers independent of the stage of the oestrous cycle, as judged by corpus luteum development. They were not associated with inflammatory cells. A total of 113 ewes (58%) had either cysts, squamous metaplasia or goblet cell hyperplasia in the cervix. A total of 148 ewes (76%) had either one of the above cervical lesions, uterine cysts, or dilated uterine glands. Acute cervicitis, generally pyogenic, was observed in 15 ewes ( 8 % ) . Chronic cervicitis, with a heavy plasma cell infiltration, was seen in 42 ewes ( 2 1% ). There was an association between the presence of cervical cysts and cervicitis (XZ1 = 16.4, P < 0.001). A fibroma about 1 cm in diameter was noted in the cervix of one ewe, and a microscopic fibroma was found in the cervix of another ewe. One ewe had a squamous cell epithelioma of the vulva.
Bacteriological Observations No strict anaerobes were isolated. Bacteria were isolated from the uterus of 59 of the 217 ewes studied ( 2 7 % ) . The genera of bacteria isolated, and in parentheses, the number of uteri from which they were isolated were as follows: Corynebacterium ( 3 1 ) , Escherichia (26), Streptococcus ( 15), Bacillus ( 7 ) , Proteus ( 3 ) , Staphylococcus ( 2 ) , Micrococcus ( 2 ) , Enterobacter ( 1 ) , Pasteurella ( 1 ) , unidentified Gram negative (3 ) , unidentified Gram positive ( 1) . Infection, in the sense of presence of bacteria within the uterus, was frequently multiple. Of the infected ewes, 36 harboured a single species of bacterium, 15 had double infections, 6 triple infections, and 2 ewes had 4 bacterial species. An association was noted between the various genera of bacteria isolated and the presence or 4ltstralian Veterinnry Journal, Vol. 5 1. July, 1975
absence of histological lesions. Because this et ul 1972). The incidence of lesions was greater paper describes observations rather than experi- in the cervix than in the uterus. The ewes would ments, it is not possible to demonstrate an aetio- not have had access to green oestrogenic pasture logical connection between bacteria isolated and for 10-18 weeks, so that the changes were more concurrent acute or chronic endometritis, but than a temporary phenomenon. 60% of ewes with E. coli and 60% of ewes with Barrett et a1 (1961) observed uterine cysts in streptococci had acute endometritis. Chronic 56.5% of ewes in selected lines from the northern endometritis was seen in 41% of ewes from tablelands of New South Wales, and Turnbull which corynebacteria was isolated, and acute et aZ (1966) recorded an incidence of 5% of endometritis in only 5%. E. coli was the organ- uterine cysts in ewes from the southern tablelands ism most frequently associated with endometritis, of New South Wales. The incidence of cysts in occurring in 12 of the 15 cases of acute endome- this study fell between these two figures. Genital tritis diagnosed histologically. Principal organ- cysts are commonly associated with clover disisms recovered from the other 3 cases included ease (Bennetts et a1 1946), but may also arise Pasteurella multocidrr, a Corynebacterium sp, and in other ways. Jubb and Kennedy (1970) sugan unidentified Gram negative organism. gest that uterine cysts may be formed in the ewe E. coli was isolated from all 4 cases of pyo- during post-partum involution. The incidence of metron studied, and in pyometron was associated this type of cyst in normal ewes is unknown, but with other bacteria, including Bacillus sp, Coryne- in the present study the high incidence of other bacterium sp, or Streptococcus sp. lesions typical of chronic oestrogen stimulation Bacteria were isolated from 16 uteri in which suggests that most of the cysts observed arose there was no endometritis, and from 15 of the from this, and oestrogenic lesions presumably due 20 uteri in which acute endometritis was diag- to clover disease are probably widespread in nosed histologically. Of the 5 cases of acute Western Australian ewes. endometritis which were bacteriologically sterile, Boyd (1969) suggested on theoretical grounds 3 occurred in the one flock, and were associated that bacteria gaining entry to the uterus at parwith embryonic mortality, The embryos appeared turition were unlikely to be an important cause to have died within four weeks of conception, of infertility in the ewe, because of the time lag and it was not possible to determine the cause between parturition and mating. However, in this of death by histology. study a significant proportion of ewes (27%) The occurrence of bacteria in the uterus was had sub-clinical infections of the uterus during associated with the presence of endometritis (Xzl the breeding season. The possibility does exist, = 23.1, P < 0.001) and also with cervicitis therefore, that bacterial infections may contribute (X21= 18.4, P < 0.001). The methods of bac- to reproductive loss in the ewe, as they do in terial isolation did not allow quantitation of the the mare (Boyd 1969) and the cow (Dawson bacteria present in the uterus, and there was no 1963), particularly since a close association was relationship between the detection of endometri- shown between the presence of bacteria and the tis, and the number of bacterial colonies isolated. occurrence of endometritis. There was a positive association between the Endometritis has been associated with chronic occurrence of bacteria in the uterus, and the oestrogen stimulation in other species, including presence of macroscopic cysts in the cervix (X21 the bitch and the cow (Fujimoto 1956). An = 20.6, P < 0.001). There also was an asso- increased incidence of pyometron has been obciation between the presence of bacteria and served in ewes with severe clover disease (Turnmacroscopic cysts in the uterus (X21- 6.7, bull et a1 1966; Fels and Neil (1968), but the P < 0.01), but no association with the presence role of subclinical endometritis in adding to the = 1.97, N.S.), clover-induced loss under commercial conditions of microscopic uterine cysts (X21 even though there was a close relationship be- has generally been overlooked. Dawson (1963) tween the presence of macroscopic and micro- found an association between endometritis and = 31.8, infertility in the cow, and Boyd (1969) has scopic cysts in the same uterus (Xzl P < 0.001). suggested that endometritis may predispose toDLscossion wards embryonic mortality. The conflicting reThere was a high incidence of lesions of the ports concerning the occurrence of embryonic type commonly associated with chronic oestro- mortality associated with clover disease (e.g., genic stimulation, including cysts, squamous Kaltenbach and Lloyd-Davies 1970; Turnbull et metaplasia and goblet cell hyperplasia in the a1 1966) may reflect variations in the incidence epithelium (Zuckerman 1940, and Hearnshaw and severity of concomitant endometritis. Australian Veterinary Journal,
Vol. 51. July, 1975
353
It i s Iikely that the cervical lesions in ewes with clover dkease facilitate entry of the vaginal flora into the uterus. Because more tracts were examined macroscopically than microscopically, the statistical association of uterine bacteria with macroscopic lesions was stronger. The strong association of uterine bacteria with cervical oestrogenic lesions, and much weaker association with uterine oestrogenic lesions, suggests that the cervix may be important in the pathogenesis of the clover disease syndrome. Work by Smith ( 1971 ) suggested that permanent functional damage of the cervix is a feature of clover disease. The occurrence of endometritis may be dependent upon the type of bacterium which gains entry. The bacterial populations in the uterus of ewes in the present study were qualitatively similar to those found in the ovine cervices by Zaki and Saber (1962) and Quinlivan (1970), except that in the uterus in this study the proportion of Gram positive bacteria was considerably lower. All Gram positive genera appeared to be proportionately reduced in incidence. In contrast, Joubert et a2 (1971) found that in the cow the Gram positive/Gram negative ratio is similar in both uterus and cervix. Furthermore, Corynebcrcteriurn pyugenes is the most important cause of pyometron in the cow (Dawson 1963). 111 the ewe, E . coli was the organism predominantly associated with pyometron. The oviductal cysts resembled those described by Donaldson ( 1 969) in the cow after prolonged treatment with oxytocin. Injection of stilboestrol into the sheep is reported to have no effect on the oviducts (Underwood et a1 1953), and as the cysts could not be associated with any lesion of clover disease in this study, their origin and significance remains obscure. The low incidence of salpingitis suggeyts it is of less important as a cause of infertility in the sheep than it is in the cow (Lombard et a1 1951). Summary
Macroscopic examination of 487 ewe reproductive tracts from IS different sources was made iit Pcrth abattoirs. Of these, 194 were examined inicrc?scopically and 2 17 bacteriologically. Observiitions were made during March to May, when ewe:. were not grazing oestrogenic pasture. Significant lesions in ovaries or oviducts were rare. Macroscopic cysts were found in the uterus o r cervix of 159 ewes ( 3 2 . 7 % ) . Microscopic lesions, including uterine or cervical cysts and squrrmcius metap1a:ia or goblet cell hyperplasia o f the epithelium, were observed in 76% of ewes studicd. Bacteria were isolated from 59 ewes, the 3 54
most common genera being Corynebacferiurn, Escherichia, Sireptococcus and Bacillus. Escherichia coli and streptococci appeared to be the most commonly associated with endometritis. Presence of bacteria was closely associated with endometritis and with cysts in the cervix. Endometritis was observed microscopically in 48 ewes (24%) and was significantly associated with the presence of cysts in the uterus or cervix. These results suggest that there is a high incidence of clover disease lesions in West Australian ewes, and that under commercial conditions clover disease may cause functional damage to the cervix, allowing passage of bacteria and resulting in endometritis. Acknowledgments
The staff and management of Western Australia Meat Exports, Robb’s Jetty, assisted greatly in collection of the specimens, and Miss Rosemary A. Heydon provided excellent technical assistance. References Barrett. J . F., Moule, G . R., Braden, A. W. H., and Harris, A. N. A . (1961) - Artst. vef. J . 37: 14. Bennetts, H. W., Underwood, E. J., and Shier, F. L. (1946) - A w l . vet. 1. 22: 2. Boyd, H. 1969) -1. Reprod. Fert. 19: Suppl. No. 8: 9. Chopping, M. H., and Lindsay, D. R. (1970) - Proc. Artsf. Sor. Atiinz. Prod. 8 : 312. Cowan, S. T., and Steel, K. J. (1965) -“Manual for the identification of Medical Bacteria”. Cambridge University Press; 1st edn, Cambridge. U.K. Dawson, F. L. M. (1963) -J . Reprod. Fert. 5: 397. Donaldson, L. E. (1969) -J . Reprod. Fert. 18: 259. Fels. H. E.. and Neil, H. G . (1968) --Aicst. J . ugric. Res. 19: 1059. Fujimoto, Y. (1956) --up. J . vet. Res. 4: 129. Hearnshaw, H., Brown, J. M., Cumming, I. A.. Goding, J . R., and Nairn, M. (1972) -1. Reprod. Fert. 2 8 : 160. Joubert, L., Khalil, 0..Bertrand, M., and Deschanel, J. P. (1971) -B~ctll. SOC. Sci. I d . M i d . romp. L ~ O T Z 73: 149. Jubb. K . V. F.. and Kennedv. P. C . (1970) -‘“Pathology of Domestic Animals”, Vol. 1 , p. 549. Academic Press; 2nd edn, New York. Kaltenbach, C . C., and Lloyd-Davies, H. (1970) Artst. 1. agric. Res. 21: 107. Liehtfoot. R. J . (1972) -J . DCD. West Artst. 13 . Aaric.. . ?Fourth Series): 102. Lombard, L., Morgan, B. B., and McNutt. S. H . (1951) - A n t . J . vet. Res. 12: 69. Quinlivan, T. D. (1970) - Artst. \vet. 1. 46: 1 I . Rossiter. R . C. (1970) - Arist. vet. J . 46: 141. Smith, I . F. (1971)-Artst. J . agric. Res. 22: 513. Turnbull. K . E.. Braden, A. W. H., and George, J. M. (1966) - Artst. 1. agric. Res. 17: 907. Underwood, F. J . , Shier, F. L., and Peterson. J. E. (2953) - Arts/. vet. J . 29: 206. Zaki, K., and Saber, M. S. (1962) -Vet. Mrd. J., Cairo. 8 : 175. (Received f o r piiblication 30 Irtly 1974) Airstraliarr I ~ e t e r i t i a r y Jourriul. Vol. 5 1. July. 1975
CSIRO, Division of Animal Health, Private Bag, P.O. Wembley, Western Australia, 6014 Mraiuction
Lamb marking rates in Western Australia are consistently below the national average, due primarily to low twinning rates, and the failure of a high proportion of ewes to bear a lamb (Chopping and Lindsay 1970). Various reasons have been proposed for this, including the nutritional conditions during mating, microbial infections of the ewe, and clover disease infertility. However, data are lacking on the importance of any of the suggested causative factors. Statistical evidence suggests that lamb marking rates are decreased in the subterranean clover areas (Lightfoot 1972), although clinical cases of clover disease with very low lambing rates and dystocia are now extremely rare throughout most of Western Australia. Although the flora of the cervix has been studied in the ewe (Zaki and Saber 1962; Quinlivan 1970), the uterine flora, and its effect on fertility in the ewe, has not been examined. This paper records the results of a pathological and microbiological abattoir survey of genitalia of ewes from commercial flocks. Materials and Methods Samples were collected between March and May in 1972 and 1973, to avoid the effects of intake of green clover. Clover pastures in Western Australia generally lose their oestrogenicity by December, when the clover dries off (Rossiter, 1970). Collections were made from groups of breeding ewes during visits on random days to the Perth abattoirs, and generally consisted of ewes culled for age. There was no selection of ewes sampled on the basis of previous fertility history. Ewes which were obviously pregnant were not examined. Where possible, the property of origin was traced, and the manager interviewed. The ewes came from properties widely dispersed throughout the 380-760 mm rainfall belt in south Western Australia, and grazed subterranean clover pastures generally typical of the district. Between 7 and 14 ewes were sampled from 18 different sources, so that a total of 194 tracts was examined macroscopically and histologically.. Of these, 145 from 15 of the sources were also e x m e d bacteriologically. An additional 293 genital tracts were collected from 4 of the 18 sources. They came from flocks of unknown reproductive history, and the additional numbers resulted from enthusiasm on the part of the abattoir workers. These tracts were examined macroscopically, an& 72 of them were also examined bacteriologically. Thus a total of 487 tracts (194 293) were examined macroscopically, and 217 (145 72) were examined bacteriologically.
+ +
Australion Veterinary Journal, Vol. 51. July, 1975
At the abattoirs, reproductive tracts were wrapped individually in paper towels and chilled on ice. Within two hours of death both uterine horns were opened aseptically in the posterior third, and a sterile bacteriological swab passed through the lumina in order to collect material for bacteriological examination. The numbers of cysts in the uterus, cervix and vagina were counted, and classified as marked (> 10 cysts), moderate (5-10 cysts), slight (< 5 cysts) or absent. Half of each ovary, portions of isthmus and ampulla from each oviduct, the mid part of each uterine hoTn, the mid part of the cervix, and the anterior vagina from each ewe were fixed in Bouin’s solution, and sections were. stained with haematoxylin and eosin. In some specimens, the vagina was not removed at the abattoirs, and could not be examined. Material from the bacteriological swabs was inoculated onto sheep blood agar plates, and the plates were incubated in a candle jar at 27°C. In addition, uterine contents from 100 of the 217 ewes were inoculated onto sheep blood agar plates which were then incubated at 37°C in an anaerobic atmosphere containing 5 % carbon dioxide. After incubation, the number of colonies was counted, and isolates were identified by the methods described by Cowan and Steel (1965). Usually the organisms were classified only by genus. Pathological and bacteriological analyses were carried out separately and the results collated at the end of the study, and analysed statistically by the Chi-square test. Results
Macroscopic Observations Massive adhesions involving the entire internal genitalia were observed in 8 ewes (1.6%). Adhesions of the ovarian bursa were seen in a further 5 ewes ( l % ) , and hydrosalpinx in 4 of these. Ovaries were judged as inactive (no corpus luteum or large follicle) in 21 ewes (4.3% 1. Most ewes with inactive ovaries came from groups with very poor body condition. Cystic endometrium was observed in ewes from 11 of the 18 groups. A total of 88 ewes ( 1 8 % ) had cystic endometrium, which was classified as marked in 30, moderate in 29, and slight in 29, Macroscopic cysts were observed in the cervix of ewes from 13 groups, with a total of 95 ewes (19.5%) af3ected. A total of 159 ewes (32.7%) had macroscopic cysts in either uterus or cervix. Pyometron was seen in 4 ewes, and hydrometron in 3 ewes. Endornetritis was observed in 3 other ewes, and in 1 of these was accompanied by embryonic resorption. Embryonic 35 1
remnants were observed in 2 other ewes. Purulent vaginitis was observed in 8 ewes, generally associated with grass seeds in the anterior vagina or posterior cervix. Unilateral aplasia of the uterine horn and caudal part of the oviduct was observed in 2 ewes, 1 on the right side, and the other on the left. One of there ewes was pregnant. One hundred and twenty-three ewes were examined for macroscopic remnants of the Wolffian duct. Parovarian cysts were found in 25 ewes (20.3%) including 1 1 on the right side, 10 on the left, and 4 bilateral. Dilated Gamer’s ducts were visible in the vagina of 16 ewes 1 3 % ) .
Microscopic Observaticns Ovary - Most ovaries appeared normal histologically. Fibrous adhesions were seen on the ovaries of those ewes in which bursa1 adhesions had been observed macroscopically. The hilar region of 5 ovaries (1.3% ) contained small cysts lined with ciliated columnar epithelium. Oviciuct - Intraepithelial cysts, often in large numbers, occurred in the ampulla of 116 ewes (61 9% ). There was no statistically significant association between oviductal intraepithelial cysts and salpingitis, endometritis or cervical cysts. Hydrosalpinx was observed in 6 ewes and was biIatera1 in 2 of these. Mild salpingitis, consisting of sniall accumulations of neutrophils, or of lymphocytes and plasma cells with epithelial degeneration, occurred in 14 ewes ( 7 % ) . In 3 ewes, there was a heavy infiltration of the oviductal tissue by eosinophils. Ulerus - Microscopic cysts were seen in the uterus of 28 ewes (14%), and dilated glands which appeared to be precursors of cysts were found in 68 ewes (35% 1. The dilated glands occurred mainly in small localized areas containing many gland buds, deep in the mucosa. Fibrosis wiis observed around blood vessels and in the muscle layers in some ewes, but periglandular fibrosis did not appear to be involved in the pathogenesis of gland dilation. Uterine glands were abundant in some ewes, and adenomyosis was observed in the uterus or cervix of 28 ewes ( I4V ) . Squamous metaplasia of the uterine epithelium was observed in one ewe. Acute endometritis was diagnosed in 29 ewes ( 15 L/r 1. in 10 of which it was pyogenic. Eschericliia coli was recovered from all ewes with pyogenic endometritis. and was frequently associated with streptococcus or corynebacteria. Chronic endometritis indicated by the presence of lymphoid nodules or large numbers of plasma cells in thc lamina propia, was observed in 37 ewes (19O’r ) , but bacteria could only be recovered 352
from 53% of these and were predominantly corynebacteria. Chronic metritis was observed in 3 ewes, and E . coli was recovered from all of these. There were associations in ewes between the presence of endometritis and uterine cysts or dilated glands (Xz1 = 13.8, P < 0.01), and between the presence of endometritis and microscopic cysts in the cervix (Xyl = 17.25, P < 0.00 1 ) . Cervix - Microscopic cysts were noted in the cervix of 70 ewes (37% ), and frequently were accompanied by an increase in glandular tissue in the cervix. Squamous metaplasia of the cervical epithelium occurred in 70 ewes, chiefly in the crypts of the posterior cervix. Hyperplastic goblet cells were seen in the cervix of 64 ewes ( 3 3 % ) , frequently associated with squamous metaplasia. In these ewes, goblet cells packed with mucus occurred in large numbers independent of the stage of the oestrous cycle, as judged by corpus luteum development. They were not associated with inflammatory cells. A total of 113 ewes (58%) had either cysts, squamous metaplasia or goblet cell hyperplasia in the cervix. A total of 148 ewes (76%) had either one of the above cervical lesions, uterine cysts, or dilated uterine glands. Acute cervicitis, generally pyogenic, was observed in 15 ewes ( 8 % ) . Chronic cervicitis, with a heavy plasma cell infiltration, was seen in 42 ewes ( 2 1% ). There was an association between the presence of cervical cysts and cervicitis (XZ1 = 16.4, P < 0.001). A fibroma about 1 cm in diameter was noted in the cervix of one ewe, and a microscopic fibroma was found in the cervix of another ewe. One ewe had a squamous cell epithelioma of the vulva.
Bacteriological Observations No strict anaerobes were isolated. Bacteria were isolated from the uterus of 59 of the 217 ewes studied ( 2 7 % ) . The genera of bacteria isolated, and in parentheses, the number of uteri from which they were isolated were as follows: Corynebacterium ( 3 1 ) , Escherichia (26), Streptococcus ( 15), Bacillus ( 7 ) , Proteus ( 3 ) , Staphylococcus ( 2 ) , Micrococcus ( 2 ) , Enterobacter ( 1 ) , Pasteurella ( 1 ) , unidentified Gram negative (3 ) , unidentified Gram positive ( 1) . Infection, in the sense of presence of bacteria within the uterus, was frequently multiple. Of the infected ewes, 36 harboured a single species of bacterium, 15 had double infections, 6 triple infections, and 2 ewes had 4 bacterial species. An association was noted between the various genera of bacteria isolated and the presence or 4ltstralian Veterinnry Journal, Vol. 5 1. July, 1975
absence of histological lesions. Because this et ul 1972). The incidence of lesions was greater paper describes observations rather than experi- in the cervix than in the uterus. The ewes would ments, it is not possible to demonstrate an aetio- not have had access to green oestrogenic pasture logical connection between bacteria isolated and for 10-18 weeks, so that the changes were more concurrent acute or chronic endometritis, but than a temporary phenomenon. 60% of ewes with E. coli and 60% of ewes with Barrett et a1 (1961) observed uterine cysts in streptococci had acute endometritis. Chronic 56.5% of ewes in selected lines from the northern endometritis was seen in 41% of ewes from tablelands of New South Wales, and Turnbull which corynebacteria was isolated, and acute et aZ (1966) recorded an incidence of 5% of endometritis in only 5%. E. coli was the organ- uterine cysts in ewes from the southern tablelands ism most frequently associated with endometritis, of New South Wales. The incidence of cysts in occurring in 12 of the 15 cases of acute endome- this study fell between these two figures. Genital tritis diagnosed histologically. Principal organ- cysts are commonly associated with clover disisms recovered from the other 3 cases included ease (Bennetts et a1 1946), but may also arise Pasteurella multocidrr, a Corynebacterium sp, and in other ways. Jubb and Kennedy (1970) sugan unidentified Gram negative organism. gest that uterine cysts may be formed in the ewe E. coli was isolated from all 4 cases of pyo- during post-partum involution. The incidence of metron studied, and in pyometron was associated this type of cyst in normal ewes is unknown, but with other bacteria, including Bacillus sp, Coryne- in the present study the high incidence of other bacterium sp, or Streptococcus sp. lesions typical of chronic oestrogen stimulation Bacteria were isolated from 16 uteri in which suggests that most of the cysts observed arose there was no endometritis, and from 15 of the from this, and oestrogenic lesions presumably due 20 uteri in which acute endometritis was diag- to clover disease are probably widespread in nosed histologically. Of the 5 cases of acute Western Australian ewes. endometritis which were bacteriologically sterile, Boyd (1969) suggested on theoretical grounds 3 occurred in the one flock, and were associated that bacteria gaining entry to the uterus at parwith embryonic mortality, The embryos appeared turition were unlikely to be an important cause to have died within four weeks of conception, of infertility in the ewe, because of the time lag and it was not possible to determine the cause between parturition and mating. However, in this of death by histology. study a significant proportion of ewes (27%) The occurrence of bacteria in the uterus was had sub-clinical infections of the uterus during associated with the presence of endometritis (Xzl the breeding season. The possibility does exist, = 23.1, P < 0.001) and also with cervicitis therefore, that bacterial infections may contribute (X21= 18.4, P < 0.001). The methods of bac- to reproductive loss in the ewe, as they do in terial isolation did not allow quantitation of the the mare (Boyd 1969) and the cow (Dawson bacteria present in the uterus, and there was no 1963), particularly since a close association was relationship between the detection of endometri- shown between the presence of bacteria and the tis, and the number of bacterial colonies isolated. occurrence of endometritis. There was a positive association between the Endometritis has been associated with chronic occurrence of bacteria in the uterus, and the oestrogen stimulation in other species, including presence of macroscopic cysts in the cervix (X21 the bitch and the cow (Fujimoto 1956). An = 20.6, P < 0.001). There also was an asso- increased incidence of pyometron has been obciation between the presence of bacteria and served in ewes with severe clover disease (Turnmacroscopic cysts in the uterus (X21- 6.7, bull et a1 1966; Fels and Neil (1968), but the P < 0.01), but no association with the presence role of subclinical endometritis in adding to the = 1.97, N.S.), clover-induced loss under commercial conditions of microscopic uterine cysts (X21 even though there was a close relationship be- has generally been overlooked. Dawson (1963) tween the presence of macroscopic and micro- found an association between endometritis and = 31.8, infertility in the cow, and Boyd (1969) has scopic cysts in the same uterus (Xzl P < 0.001). suggested that endometritis may predispose toDLscossion wards embryonic mortality. The conflicting reThere was a high incidence of lesions of the ports concerning the occurrence of embryonic type commonly associated with chronic oestro- mortality associated with clover disease (e.g., genic stimulation, including cysts, squamous Kaltenbach and Lloyd-Davies 1970; Turnbull et metaplasia and goblet cell hyperplasia in the a1 1966) may reflect variations in the incidence epithelium (Zuckerman 1940, and Hearnshaw and severity of concomitant endometritis. Australian Veterinary Journal,
Vol. 51. July, 1975
353
It i s Iikely that the cervical lesions in ewes with clover dkease facilitate entry of the vaginal flora into the uterus. Because more tracts were examined macroscopically than microscopically, the statistical association of uterine bacteria with macroscopic lesions was stronger. The strong association of uterine bacteria with cervical oestrogenic lesions, and much weaker association with uterine oestrogenic lesions, suggests that the cervix may be important in the pathogenesis of the clover disease syndrome. Work by Smith ( 1971 ) suggested that permanent functional damage of the cervix is a feature of clover disease. The occurrence of endometritis may be dependent upon the type of bacterium which gains entry. The bacterial populations in the uterus of ewes in the present study were qualitatively similar to those found in the ovine cervices by Zaki and Saber (1962) and Quinlivan (1970), except that in the uterus in this study the proportion of Gram positive bacteria was considerably lower. All Gram positive genera appeared to be proportionately reduced in incidence. In contrast, Joubert et a2 (1971) found that in the cow the Gram positive/Gram negative ratio is similar in both uterus and cervix. Furthermore, Corynebcrcteriurn pyugenes is the most important cause of pyometron in the cow (Dawson 1963). 111 the ewe, E . coli was the organism predominantly associated with pyometron. The oviductal cysts resembled those described by Donaldson ( 1 969) in the cow after prolonged treatment with oxytocin. Injection of stilboestrol into the sheep is reported to have no effect on the oviducts (Underwood et a1 1953), and as the cysts could not be associated with any lesion of clover disease in this study, their origin and significance remains obscure. The low incidence of salpingitis suggeyts it is of less important as a cause of infertility in the sheep than it is in the cow (Lombard et a1 1951). Summary
Macroscopic examination of 487 ewe reproductive tracts from IS different sources was made iit Pcrth abattoirs. Of these, 194 were examined inicrc?scopically and 2 17 bacteriologically. Observiitions were made during March to May, when ewe:. were not grazing oestrogenic pasture. Significant lesions in ovaries or oviducts were rare. Macroscopic cysts were found in the uterus o r cervix of 159 ewes ( 3 2 . 7 % ) . Microscopic lesions, including uterine or cervical cysts and squrrmcius metap1a:ia or goblet cell hyperplasia o f the epithelium, were observed in 76% of ewes studicd. Bacteria were isolated from 59 ewes, the 3 54
most common genera being Corynebacferiurn, Escherichia, Sireptococcus and Bacillus. Escherichia coli and streptococci appeared to be the most commonly associated with endometritis. Presence of bacteria was closely associated with endometritis and with cysts in the cervix. Endometritis was observed microscopically in 48 ewes (24%) and was significantly associated with the presence of cysts in the uterus or cervix. These results suggest that there is a high incidence of clover disease lesions in West Australian ewes, and that under commercial conditions clover disease may cause functional damage to the cervix, allowing passage of bacteria and resulting in endometritis. Acknowledgments
The staff and management of Western Australia Meat Exports, Robb’s Jetty, assisted greatly in collection of the specimens, and Miss Rosemary A. Heydon provided excellent technical assistance. References Barrett. J . F., Moule, G . R., Braden, A. W. H., and Harris, A. N. A . (1961) - Artst. vef. J . 37: 14. Bennetts, H. W., Underwood, E. J., and Shier, F. L. (1946) - A w l . vet. 1. 22: 2. Boyd, H. 1969) -1. Reprod. Fert. 19: Suppl. No. 8: 9. Chopping, M. H., and Lindsay, D. R. (1970) - Proc. Artsf. Sor. Atiinz. Prod. 8 : 312. Cowan, S. T., and Steel, K. J. (1965) -“Manual for the identification of Medical Bacteria”. Cambridge University Press; 1st edn, Cambridge. U.K. Dawson, F. L. M. (1963) -J . Reprod. Fert. 5: 397. Donaldson, L. E. (1969) -J . Reprod. Fert. 18: 259. Fels. H. E.. and Neil, H. G . (1968) --Aicst. J . ugric. Res. 19: 1059. Fujimoto, Y. (1956) --up. J . vet. Res. 4: 129. Hearnshaw, H., Brown, J. M., Cumming, I. A.. Goding, J . R., and Nairn, M. (1972) -1. Reprod. Fert. 2 8 : 160. Joubert, L., Khalil, 0..Bertrand, M., and Deschanel, J. P. (1971) -B~ctll. SOC. Sci. I d . M i d . romp. L ~ O T Z 73: 149. Jubb. K . V. F.. and Kennedv. P. C . (1970) -‘“Pathology of Domestic Animals”, Vol. 1 , p. 549. Academic Press; 2nd edn, New York. Kaltenbach, C . C., and Lloyd-Davies, H. (1970) Artst. 1. agric. Res. 21: 107. Liehtfoot. R. J . (1972) -J . DCD. West Artst. 13 . Aaric.. . ?Fourth Series): 102. Lombard, L., Morgan, B. B., and McNutt. S. H . (1951) - A n t . J . vet. Res. 12: 69. Quinlivan, T. D. (1970) - Artst. \vet. 1. 46: 1 I . Rossiter. R . C. (1970) - Arist. vet. J . 46: 141. Smith, I . F. (1971)-Artst. J . agric. Res. 22: 513. Turnbull. K . E.. Braden, A. W. H., and George, J. M. (1966) - Artst. 1. agric. Res. 17: 907. Underwood, F. J . , Shier, F. L., and Peterson. J. E. (2953) - Arts/. vet. J . 29: 206. Zaki, K., and Saber, M. S. (1962) -Vet. Mrd. J., Cairo. 8 : 175. (Received f o r piiblication 30 Irtly 1974) Airstraliarr I ~ e t e r i t i a r y Jourriul. Vol. 5 1. July. 1975
