- Email: [email protected]
pyometra complex in the bitch
Theriogenology 66 (2006) 1530–1536 www.journals.elsevierhealth.com/periodicals/the
A model for cystic endometrial hyperplasia/pyometra complex in the bitch N. Arora *, J. Sandford, G.F. Browning, J.R. Sandy, P.J. Wright Department of Veterinary Science, The University of Melbourne, 250 Princes Highway, Werribee 3030, Vic., Australia
Abstract The objective of this study was to develop a reliable model for the study of the cystic endometrial hyperplasia and pyometra complex (CEH/P) in the bitch. Greyhound bitches (n = 15) were ovariectomised and allocated into three groups (Group 1, n = 5; Group 2, n = 5; Group 3, n = 10, including 5 used from Group 1). Simulated proestrus, estrus and diestrus were induced by treatment with estradiol benzoate and megestrol acetate. The duration of cervical opening during estrus was determined by the intra-vaginal infusion of radio-opaque medium and subsequent radiography of the uterus (Group 1). One milliliter of a culture of Escherichia coli (with five uro-pathogenic virulence factors as identified by PCR: pap, sfa, hlyA, cnf1 and fim) was inoculated intra-vaginally daily throughout the simulated estrus (Group 2). One milliliter of the culture (n = 6) or sterile Luria–Bertani broth (n = 4) was introduced directly into the uterus on simulated diestrus Days 8 or 12 (Group 3). Necropsies were performed 12 and 7–14 days after the inoculation (Groups 2 and 3). The cervix remained open throughout the duration of simulated estrus (5–6 days) in four out of five bitches, and for a shorter duration (3 days of a 6-day estrus period) in one bitch (Group 1). CEH/P was induced by inoculation of bacteria into the uterus (10/10 bitches) but not into the vagina (0/5 bitches), (P = 0.003). A model for the study of CEH/P has been validated. # 2006 Elsevier Inc. All rights reserved. Keywords: Cystic endometrial hyperplasia; Pyometra; Escherichia coli; Canine; Dog
1. Introduction The cystic endometrial hyperplasia/pyometra complex (CEH/P) is an acute or chronic, poly-systemic, diestrual disorder of the adult, ovary-intact bitch characterised by hyperplasia of the endometrium and infiltration of inflammatory cells, which may be present in all layers of the uterus [1–3]. CEH/P results from an infection, commonly Escherichia coli (E. coli), ascending from the vagina [4]. In previous studies, CEH was induced in intact bitches by the administration of high doses of estrogen
* Corresponding author. Tel.: +61 3 9731 2333; fax: +61 3 9731 2366. E-mail address: [email protected] (N. Arora). 0093-691X/$ – see front matter # 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.theriogenology.2006.02.019
and progesterone for prolonged intervals, by uterine biopsies, uterine scarification, and uterine irritants such as silk suture material [5–7]. Furthermore, CEH/P can be induced in ovary-intact bitches by the intra-uterine inoculation of E. coli during diestrus, with or without ligation of cervix [5,8]. A model for CEH using ovariectomised bitches treated with a uterine irritant (silk suture) and estradiol benzoate and megestrol acetate has been described [9,10]. However, no model for CEH/P involving inoculation of bacteria into the vagina or the uterus of ovariectomised bitches has been established. The objective of this study was to validate such a model in order to facilitate further studies of the pathogenesis, treatment and control of this disease. A preliminary study was undertaken to determine whether E. coli, present in an inoculum, would enter the uterine lumen
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
1531
after being infused intra-vaginally in bitches during a simulated estrus.
diestrus (Di1) was the day when SCI decreased by at least 20% [12].
2. Materials and methods
2.3. Plasma progesterone assay
The study comprised three experiments. The objectives were to determine: (i) the duration of cervical patency in ovariectomised bitches during simulated estrus; (ii) the effect of intra-vaginal inoculation of E. coli on the occurrence of CEH/P; (iii) the effect of intra-uterine inoculation of E. coli on the occurrence of CEH/P.
Blood samples were collected by jugular venipuncture before ovariectomy. Plasma progesterone concentrations were determined in a commercially available coated tube radioimmunoassay (Coat-A-Count Progesterone; Diagnostic Products Corporation, Los Angeles, CA, USA) [13]. All samples were measured over three assays; the sensitivities were 0.03, 0.03 and 0.07 ng/mL, respectively.
2.1. Animals and treatments Mature, ovary-intact greyhound bitches (n = 15) aged 2–3 years, and in anestrus (determined by physical examination of external genitalia, vaginal cytology and plasma progesterone concentrations) were ovariectomised. The bitches were allocated into three groups: Group 1, n = 5; Group 2, n = 5; Group 3, n = 10 (including five bitches from Group 1). Ovariectomies were performed under standard anesthesia protocols. Anesthesia was induced using propofol (Rapinovet; Schering-Plough, North Ryde, Australia) and maintained using isoflurane (Isoflurane; Advanced Anaesthesia Specialists, North Ryde, Australia). Two weeks after ovariectomy, simulated stages of the estrous cycle were induced by treating the bitches with estradiol benzoate (Mesalin; Intervet, Bendigo East, Australia; 0.6–4.8 mg/kg im, twice daily for 13 days), followed by megestrol acetate (Ovarid; Jurox Pty Ltd., Rutherford, Australia; 2.0 mg/kg orally, once a day, until euthanasia) [10]. 2.2. Vaginal cytology, stages of the cycle Vaginal smears were taken before ovariectomy (to confirm anestrus) and during the period of treatment with estradiol benzoate and megestrol acetate (to determine the simulated stages of the estrous cycle). The smears were taken every 3 days, from Days 1–7 of treatment, then every 2 days from Day 8 of treatment until Day 4 of estrus, and then daily from Day 5 of estrus until Day 1 of diestrus. The vaginal smears were stained with Diff-Quick (Laboratory Aids, Narrabeen, Australia) and superficial cell indices (SCI) determined [11]. Anestrus was when SCI was 0% and plasma progesterone concentrations were <0.5 ng/mL. Estrus was the period when the SCI was >90% and the bitches showed positive postural reflexes (marked tail flagging, vulva lifting in response to perineal stroking). Day 1 of
2.4. Preparation of E. coli inoculum A strain of Escherichia coli (E. coli) isolated from a naturally occurring case of pyometra and possessing genes for five uro-pathogenic virulence factors (pap, sfa, hlyA, cnf1 and fim) as determined by polymerase chain reaction was used [14]. A loop full of E. coli seed culture was used to inoculate 10 mL of Luria–Bertani broth and the culture was incubated in a shaking incubator at 37 8C at 200 rpm for 22 h. The concentration of bacteria in the culture was determined using a standard plate count method and compared to the optical density of 10 1 and 10 2 dilutions of the culture at 595 nm (A595). The A595 measurements were used to estimate the number of bacteria in the culture immediately prior to inoculation. 2.5. Necropsy The bitches were euthanased by intravenous injection of pentobarbitone sodium (Lethabarb; Virbac, Peakhurst, Australia). A ventral midline incision was made, a clamp was placed on the anterior vagina and the genital tract was removed. 2.6. Uterine microbiology Samples were collected from the uterus for microbiological culture. When there was no accumulation of fluid in the uterus, 5 mL of saline was flushed through the uterine horns and collected from the body of the uterus. The uterine contents were serially diluted (1–9 mL) up to 10 7 with saline, and 100 mL of each dilution was inoculated on to Luria–Bertani agar plates. After 24 h of incubation, the number of colonies on each plate was counted to determine the concentration of E. coli in the uterine contents.
1532
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
2.7. Uterine histology Transverse and longitudinal sections from the middle of each uterine horn and the body of the uterus were taken and fixed in Bouin’s solution for 24 h, followed by 70% alcohol, then embedded in paraffin and sectioned using standard procedures. The sections were stained with haematoxylin and eosin. 2.8. Statistical analysis The bitches were randomly allocated into the test and control groups using the lottery method. Fisher’s exact test and the ‘‘t’’ test were applied to data as appropriate using a computer program ‘‘STAT-SAK’’ (The Statistician’s Swiss Army Knife, Version 2.40, by Gerard E. Dallal Malden, MA 02148). 2.9. Duration of cervical patency (experiment 1) The passage of radio-opaque medium from the vagina to the uterus was determined each day during simulated estrus until cervical closure. The dogs were anesthetised as described previously. Sixty milliliters of a 1 in 4 solution of ‘Isovue-370’ (Bracco, Regional Health Care Products Group, NSW, Australia) was infused into the genital tract [15] using an intra-vaginal balloon catheter (Lyppard, Cheltenham, Australia) inflated with 50 mL of air. The catheter was left in place for 10 min. A left lateral radiograph was taken before infusion and immediately after infusion, followed by left lateral and ventro-dorsal radiographs 10 min later. These bitches received intra-uterine inoculation with E. coli or with control broth during their subsequent simulated diestrus. 2.10. Intra-vaginal inoculation of E. coli and incidence of CEH/P (experiment 2) One milliliter of the E. coli culture (109–1010colony forming units (cfu)) was inoculated intra-vaginally in each bitch (n = 5) daily throughout the simulated estrus. The hindquarters of each bitch were raised during the inoculation and for 5 min afterwards. The general health and feed intake of each dog was monitored throughout the study. Blood samples for hematology were taken daily prior to the inoculation of the bacteria and on the day of euthanasia. Ultrasonographic examination of the uterus of each bitch was performed every 4 days after the last inoculation. Necropsies were performed 12 days after the last inoculation.
2.11. Intra-uterine inoculation of E. coli and incidence of CEH/P (experiment 3) Ten bitches were allocated into two groups of five. Each group comprised three test and two control dogs. One milliliter of E. coli culture (109–1010cfu) or sterile broth was inoculated (0.5 mL into each uterine horn) on Day 12 (first group) or Day 8 (second group) of simulated diestrus by laparotomy using a 24 gauge intra-venous catheter (Optiva, Johnson & Johnson Jntl, Belgium). The inoculations were performed under the anesthetic regimen described previously. Feed intake, disposition and demeanour were monitored throughout the study in both the groups. In addition, daily water intake was measured after the inoculation in the second group. Blood samples for hematology (collected by jugular venipuncture) were taken prior to inoculation and then every 2 or 3 days after inoculation and on the day of euthanasia. Vaginal smears were taken every 3 days after inoculation (second group) and on the day of euthanasia (both groups). Ultrasonographic examination of uterus was performed on the day of euthanasia (one bitch) and on a day before euthanasia (nine bitches). Necropsies were initially scheduled 14 days after inoculation, or before then if any of the bitches showed any signs of ill health. Necropsies were performed 14 days (first group), 7 days (1/5 bitch, second group) and 9 days (4/5 bitches, second group) after inoculation. 3. Results All the bitches were in anestrus at the commencement of the study as determined by small vulvar size, vaginal cytology (SCI 0%) and plasma progesterone concentrations (<0.5 ng/mL). The vaginal superficial cell index was >50% on Day 7 of the estradiol benzoate treatment and increased to >90% by Day 13 of treatment. Estrus commenced 10–11 days after the start of treatment. 3.1. Duration of cervical patency (experiment 1) Radiographs taken before, immediately after the infusion and 10 min later, demonstrated the entry of radio-opaque fluid into the uterine horns through the cervix. In four of five bitches, the cervix remained open throughout simulated estrus. In one bitch, the cervix closed on Days 3 or 4 of a simulated estrus of 6 days. The degree of patency of the cervix reflected by the amount of fluid in the uterus varied (Table 1, Fig. 1a–c); however, this variation was not related to the stage of estrus.
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
1533
Table 1 The presence of radio-opaque fluid in the uterus 10 min after intra-vaginal infusion in dogs Dog ID
NA6 NA7 NA8 NA9 NA10
Duration of estrus (d) 6 6 5 6 6
Day of estrus 1 + ua u u ++
Day of diestrus 2 ++ + ++ + ++
3 ++ u u u ++
4 + 0 + ++ ++
5 ++ 0 ++ + ++
6 ++ nd ++ + ++
1
4/5 b
nd nd ++ nd nd
0 0 0 0 0
0, nil; +, small amount; ++, large amount. a Unclear. b Not done.
3.2. Intra-vaginal inoculation of bacteria and incidence of CEH/P (experiment 2) All the dogs were alert and bright and showed no change in feed intake throughout the period of
observation. There were no significant differences in leucocyte concentrations in blood samples taken during the period of inoculations and on the day of euthanasia. Ultrasonographic examination of the uterus in each bitch revealed no abnormality. No gross
Fig. 1. (a) Bitch at Day 4 of diestrus. Note that the radio-opaque fluid did not pass the cervix (shown by an arrow) indicating a closed cervix (graded as 0). (b) Bitch at Day 4 of estrus. Note that a small amount of radio-opaque fluid (shown by an arrow) passed through the cervix into the uterine horns (graded as +). (c) Bitch at Day 2 of estrus. Note that a large amount of radio-opaque fluid (shown by an arrow) passed through the cervix into the uterine horns (graded as ++).
1534
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
Fig. 2. (a) The uterus of a bitch 12 days after the daily intra-vaginal inoculation with E. coli during simulated estrus (Group 2). The uterus is small and contains no fluid. There is no evidence of CEH/P. (b) Light micrograph of a transverse section of the uterus of a bitch 12 days after daily intra-vaginal inoculation with E. coli during simulated estrus (Group 2). No signs of inflammatory response or cysts. There is no evidence of CEH/P (40 hematoxylin eosin stain).
pathological changes were observed in the uteri at necropsy (Fig. 2a). The uterine flushings obtained were clear, colourless and watery in consistency. E. coli were not isolated from the uterine flushings. Histologically, the tissue sections showed no signs of CEH/P (Fig. 2b). 3.3. Intra-uterine inoculation of bacteria and incidence of CEH/P (experiment 3) A small reduction in feed intake, slight purulent vaginal discharge and leukocytes in vaginal smears were observed in all the test dogs of first group. The total white blood cell numbers were higher in all the test dogs than in control dogs (P > 0.05) on the day of euthanasia but still remained within the reference range (3.6– 13.3 109 L 1), except one dog where the numbers raised above the reference range (15.6 109 L 1).
Fig. 3. (a) The uterus of a test bitch 9 days after intra-uterine inoculation with E. coli during simulated diestrus (Group 3). Note that both the uterine horns were enlarged and filled with fluid, indicating CEH/P. (b) The uterus of a control bitch 9 days after intra-uterine inoculation with Luria–Bertani broth during simulated diestrus (Group 3). The uterus is small and contains no fluid. There is no evidence of CEH/P.
These animals were alert and bright throughout. In the test dogs of the second group, there were no changes in feed intake, but there was increased water intake. No vaginal discharge and no infiltration of leukocytes in vaginal smears were observed. The total white blood cell numbers were higher in the test dogs than in control dogs (P > 0.05) on the day of euthanasia but remained within the reference range (3.6–13.3 109 L 1). The control dogs of both the groups showed no changes in any of these characteristics. One test dog of the second group became depressed on Day 7 after inoculation. Ultrasound examination revealed a fluid-filled uterus. The dog was euthanized the same day. Other dogs of the second group who remained bright and alert throughout were
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
euthanased the following day. Ultrasonographic examination of the uterus of test dogs of both the groups indicated large, fluid filled uteri. At necropsy, the uteri of test dogs were filled with purulent exudate (measured in second group dogs) of yellow to reddish brown colour and a fetid odour whereas the uteri obtained from the control dogs were grossly normal (Fig. 3a and b). Colonies of E. coli were isolated from the uterine contents of the test dogs and the concentration of E. coli in each milliliter of the purulent exudate was calculated in for each dog (Table 2). The uterine flushings obtained from the control dogs did not produce E. coli colonies on the agar plates. Histological examination of each uterine horn and the body of the uterus revealed large cystic endometrial glands and increased endometrial thickness, an influx of inflammatory cells into the glandular
1535
Table 2 The volume of purulent exudate in the canine uterus and the number of E. coli calculated in each milliliter of purulent exudates Dog ID
T/C a
E. coli colonies (cfub/mL)
Purulent exudate (mL)
NA6 NA7 NA8 NA9 NA10 NA12 NA13 NA14 NA15 NA16
T T T C C T C T C T
1.7 109 1.77 109 2.4 109 Nil Nil 5.6 109 Nil 1.19 1010 Nil 1.49 109
ndc nd nd nd nd 25–26 Nil 15–16 Nil 35
a b c
Test/control. Colony forming units. Not done.
and uterine luminal areas (Fig. 4a) and bacterial colonies. Histological examination of uterine tissues from the control dogs did not reveal any changes in the endometrial glands or any infiltration of inflammatory cells (Fig. 4b). 4. Discussion
Fig. 4. (a) Light micrograph of a transverse section of the uterus of a test bitch 14 days after intra-uterine inoculation with E. coli during simulated diestrus (Group 3). There is severe inflammation with cystic glands (~) containing inflammatory debris and the lumen (*) is filled with inflammatory cells. (40 hematoxylin eosin stain). (b) Light micrograph of a transverse section of the uterus of a control bitch 14 days after intra-uterine inoculation with Luria–Bertani broth during simulated diestrus (Group 3). There is no evidence of CEH/P (40 hematoxylin eosin stain).
The results of this study indicate that intra-uterine inoculation of E. coli during simulated diestrus in ovariectomised bitches can reliably induce CEH/P. This validates the ovariectomised bitch model for CEH/P and was consistent with previous findings of the development of CEH/P resulting from the intra-uterine inoculation of E. coli during diestrus [8,16,17]. The occurrence of CEH/P in the estradiol-progestagen treated ovariectomised bitch model adds further confirmation to the validity of this model for the study of the estrous cycle [10]. The failure of intra-vaginal inoculation to induce CEH/P is similar to previous findings where intravaginal inoculation of E. coli during naturally occurring estrus also failed to induce CEH/P [18]. That study involved only one intra-vaginal inoculation during each estrus; in the present study, daily inoculations of pathogenic E. coli during simulated estrus were performed. The lack of CEH/P in spite of the greater frequency of inoculation with a pathogenic strain of E. coli indicated that the normal, healthy uterus during estrus (simulated or natural) has strong defences and can eliminate large doses of pathogenic E. coli [5]. It also suggests that some other factor(s) may be required to initiate infection and assist the persistence and action of bacteria in the uterus [18].
1536
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
The cervix remained open throughout the simulated estrus in four of five bitches in our model, in contrast to a previous study in ovary-intact bitches where the cervix closed 2.6 1 days before Day 1 diestrus [15]. The methodologies of both the studies involved similar amounts of fluid inoculated into the vagina under similar pressure. The reason for the different finding is unclear, but suggests that bitches in simulated estrus should be more susceptible to uterine infections than bitches in natural estrus because of the increased time and opportunity for bacteria to enter the uterus. The finding of a purulent vaginal discharge in the test dogs of the first group receiving intra-uterine inoculation indicated an open-cervix pyometra, whereas in the second group, the lack of such findings indicated a closed cervix pyometra. This difference may have been due to the fact that the dogs used in first group had previously been used for the part of the study examining the patency of the cervix by performing daily infusions of radio-opaque fluid. This treatment may have affected cervical function. In conclusion, a model for cystic endometrial hyperplasia and pyometra using ovariectomised bitches treated with estradiol benzoate, megestrol acetate and E. coli has been validated. This model can be easily and reliably reproduced and used as a basis for future studies of the pathogenesis and prevention of the disease. Future studies to refine the model will involve reduction of the number of E. coli in the inoculum, use of E. coli without the UVF in early diestrus and intra-uterine inoculations during estrus. Acknowledgments We thank Dr. Ben Landon for his guidance in surgical technique, Dr. Cathy Beck and Sylvia Meekings for their assistance in radiology, Dr. Jenny Charles for histological advice and Fiona Armour, Joanne Allen and Margaret Sung for their technical support. The funding support of the Victorian Canine Association, Canine Research Foundation is gratefully acknowledged. References [1] Dow C. The cystic hyperplasia–pyometra complex in the bitch. Vet Rec 1957;69:1409–15.
[2] Hardy RM, Osborne CA. Canine pyometra: pathophysiology, diagnosis and treatment of uterine and extra-uterine lesions. J Am Anim Hosp Assoc 1974;10:245–68. [3] Kennedy PC, Miller RB. The female genital system. In: Palmer N, editor. Pathology of domestic animals. Academic Press; 1993. p. 349–86. [4] Johnston SD, Kustritz MVR, Olson PNS. Disorders of the canine uterus and uterine tubes (oviducts). In: Kersey R, editor. Canine and feline theriogenology. W.B. Saunders Company; 2001. p. 206–24. [5] Noakes DE, Dhaliwal GK, England GCW. Cystic endometrial hyperplasia/pyometra in dogs: a review of the causes and pathogenesis. J Reprod Fertil Suppl 2001;57:395–406. [6] Nomura K. Induction of a deciduoma in the dog. J Vet Med Sci 1994;56:365–9. [7] Bosschere HD, Ducatelle R, Tshamala M. Is mechanically induced cystic endometrial hyperplasia (CEH) a suitable model for study of spontaneously occurring CEH in the uterus of the bitch? Reprod Dom Anim 2002;37:152–7. [8] Bosschere HD, Ducatelle R, Tshamala M. Uterine oestrogen and progesterone receptor expression in experimental pyometra in the bitch. J Comp Pathol 2003;128:99–106. [9] Chen YMM, Wright PJ, Lee CS. A model for the study of cystic endometrial hyperplasia in bitches. J Reprod Fertil Suppl 2001;57:407–14. [10] Chu PY, Lee CS, Moore PF, Wright PJ. Oestrogen and progesterone treated ovariectomised bitches: a model for the study of uterine function. J Reprod Fertil Suppl 2001;57:45–54. [11] Schutte AP. Canine vaginal cytology. III. Compilation and evaluation of cellular indices. J Small Anim Pract 1967;8: 313–7. [12] Holst P, Phemister R. Onset of diestrus in the Beagle bitch: definition and significance. Am J Vet Res 1974;35:401–6. [13] Wright PJ. Practical aspects of the estimation of the time of ovulation and of insemination in the bitch. Aust Vet J 1991;68: 10–3. [14] Chen YMM, Wright PJ, Lee CS, Browning GF. Uropathogenic virulence factors in isolates of Escherichia coli from clinical cases of canine pyometra and feces of healthy bitches. Vet Microbiol 2003;94:57–69. [15] Silva LDM, Onclin K, Verstegen JP. Cervical opening in relation to progesterone and oestradiol during heat in beagle bitches. J Reprod Fertil 1995;104:85–90. [16] Nomura K. Canine pyometra with cystic endometrial hyperplasia experimentally induced by E. coli inoculation. J Vet Sci 1983;45:237–40. [17] Nomura K, Funahashi H. Histological characteristics of canine deciduoma induced by intrauterine inoculation of E. coli suspension. J Vet Med Sci 1999;61:433–8. [18] Nomura K, Yoshida K, Funahashi H, Shimada Y. The possibilities of uterine invasion of Escherichia coli inoculated into the vagina and development of endometritis in the bitches. J Anim Reprod 1988;34:199–203.
A model for cystic endometrial hyperplasia/pyometra complex in the bitch N. Arora *, J. Sandford, G.F. Browning, J.R. Sandy, P.J. Wright Department of Veterinary Science, The University of Melbourne, 250 Princes Highway, Werribee 3030, Vic., Australia
Abstract The objective of this study was to develop a reliable model for the study of the cystic endometrial hyperplasia and pyometra complex (CEH/P) in the bitch. Greyhound bitches (n = 15) were ovariectomised and allocated into three groups (Group 1, n = 5; Group 2, n = 5; Group 3, n = 10, including 5 used from Group 1). Simulated proestrus, estrus and diestrus were induced by treatment with estradiol benzoate and megestrol acetate. The duration of cervical opening during estrus was determined by the intra-vaginal infusion of radio-opaque medium and subsequent radiography of the uterus (Group 1). One milliliter of a culture of Escherichia coli (with five uro-pathogenic virulence factors as identified by PCR: pap, sfa, hlyA, cnf1 and fim) was inoculated intra-vaginally daily throughout the simulated estrus (Group 2). One milliliter of the culture (n = 6) or sterile Luria–Bertani broth (n = 4) was introduced directly into the uterus on simulated diestrus Days 8 or 12 (Group 3). Necropsies were performed 12 and 7–14 days after the inoculation (Groups 2 and 3). The cervix remained open throughout the duration of simulated estrus (5–6 days) in four out of five bitches, and for a shorter duration (3 days of a 6-day estrus period) in one bitch (Group 1). CEH/P was induced by inoculation of bacteria into the uterus (10/10 bitches) but not into the vagina (0/5 bitches), (P = 0.003). A model for the study of CEH/P has been validated. # 2006 Elsevier Inc. All rights reserved. Keywords: Cystic endometrial hyperplasia; Pyometra; Escherichia coli; Canine; Dog
1. Introduction The cystic endometrial hyperplasia/pyometra complex (CEH/P) is an acute or chronic, poly-systemic, diestrual disorder of the adult, ovary-intact bitch characterised by hyperplasia of the endometrium and infiltration of inflammatory cells, which may be present in all layers of the uterus [1–3]. CEH/P results from an infection, commonly Escherichia coli (E. coli), ascending from the vagina [4]. In previous studies, CEH was induced in intact bitches by the administration of high doses of estrogen
* Corresponding author. Tel.: +61 3 9731 2333; fax: +61 3 9731 2366. E-mail address: [email protected] (N. Arora). 0093-691X/$ – see front matter # 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.theriogenology.2006.02.019
and progesterone for prolonged intervals, by uterine biopsies, uterine scarification, and uterine irritants such as silk suture material [5–7]. Furthermore, CEH/P can be induced in ovary-intact bitches by the intra-uterine inoculation of E. coli during diestrus, with or without ligation of cervix [5,8]. A model for CEH using ovariectomised bitches treated with a uterine irritant (silk suture) and estradiol benzoate and megestrol acetate has been described [9,10]. However, no model for CEH/P involving inoculation of bacteria into the vagina or the uterus of ovariectomised bitches has been established. The objective of this study was to validate such a model in order to facilitate further studies of the pathogenesis, treatment and control of this disease. A preliminary study was undertaken to determine whether E. coli, present in an inoculum, would enter the uterine lumen
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
1531
after being infused intra-vaginally in bitches during a simulated estrus.
diestrus (Di1) was the day when SCI decreased by at least 20% [12].
2. Materials and methods
2.3. Plasma progesterone assay
The study comprised three experiments. The objectives were to determine: (i) the duration of cervical patency in ovariectomised bitches during simulated estrus; (ii) the effect of intra-vaginal inoculation of E. coli on the occurrence of CEH/P; (iii) the effect of intra-uterine inoculation of E. coli on the occurrence of CEH/P.
Blood samples were collected by jugular venipuncture before ovariectomy. Plasma progesterone concentrations were determined in a commercially available coated tube radioimmunoassay (Coat-A-Count Progesterone; Diagnostic Products Corporation, Los Angeles, CA, USA) [13]. All samples were measured over three assays; the sensitivities were 0.03, 0.03 and 0.07 ng/mL, respectively.
2.1. Animals and treatments Mature, ovary-intact greyhound bitches (n = 15) aged 2–3 years, and in anestrus (determined by physical examination of external genitalia, vaginal cytology and plasma progesterone concentrations) were ovariectomised. The bitches were allocated into three groups: Group 1, n = 5; Group 2, n = 5; Group 3, n = 10 (including five bitches from Group 1). Ovariectomies were performed under standard anesthesia protocols. Anesthesia was induced using propofol (Rapinovet; Schering-Plough, North Ryde, Australia) and maintained using isoflurane (Isoflurane; Advanced Anaesthesia Specialists, North Ryde, Australia). Two weeks after ovariectomy, simulated stages of the estrous cycle were induced by treating the bitches with estradiol benzoate (Mesalin; Intervet, Bendigo East, Australia; 0.6–4.8 mg/kg im, twice daily for 13 days), followed by megestrol acetate (Ovarid; Jurox Pty Ltd., Rutherford, Australia; 2.0 mg/kg orally, once a day, until euthanasia) [10]. 2.2. Vaginal cytology, stages of the cycle Vaginal smears were taken before ovariectomy (to confirm anestrus) and during the period of treatment with estradiol benzoate and megestrol acetate (to determine the simulated stages of the estrous cycle). The smears were taken every 3 days, from Days 1–7 of treatment, then every 2 days from Day 8 of treatment until Day 4 of estrus, and then daily from Day 5 of estrus until Day 1 of diestrus. The vaginal smears were stained with Diff-Quick (Laboratory Aids, Narrabeen, Australia) and superficial cell indices (SCI) determined [11]. Anestrus was when SCI was 0% and plasma progesterone concentrations were <0.5 ng/mL. Estrus was the period when the SCI was >90% and the bitches showed positive postural reflexes (marked tail flagging, vulva lifting in response to perineal stroking). Day 1 of
2.4. Preparation of E. coli inoculum A strain of Escherichia coli (E. coli) isolated from a naturally occurring case of pyometra and possessing genes for five uro-pathogenic virulence factors (pap, sfa, hlyA, cnf1 and fim) as determined by polymerase chain reaction was used [14]. A loop full of E. coli seed culture was used to inoculate 10 mL of Luria–Bertani broth and the culture was incubated in a shaking incubator at 37 8C at 200 rpm for 22 h. The concentration of bacteria in the culture was determined using a standard plate count method and compared to the optical density of 10 1 and 10 2 dilutions of the culture at 595 nm (A595). The A595 measurements were used to estimate the number of bacteria in the culture immediately prior to inoculation. 2.5. Necropsy The bitches were euthanased by intravenous injection of pentobarbitone sodium (Lethabarb; Virbac, Peakhurst, Australia). A ventral midline incision was made, a clamp was placed on the anterior vagina and the genital tract was removed. 2.6. Uterine microbiology Samples were collected from the uterus for microbiological culture. When there was no accumulation of fluid in the uterus, 5 mL of saline was flushed through the uterine horns and collected from the body of the uterus. The uterine contents were serially diluted (1–9 mL) up to 10 7 with saline, and 100 mL of each dilution was inoculated on to Luria–Bertani agar plates. After 24 h of incubation, the number of colonies on each plate was counted to determine the concentration of E. coli in the uterine contents.
1532
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
2.7. Uterine histology Transverse and longitudinal sections from the middle of each uterine horn and the body of the uterus were taken and fixed in Bouin’s solution for 24 h, followed by 70% alcohol, then embedded in paraffin and sectioned using standard procedures. The sections were stained with haematoxylin and eosin. 2.8. Statistical analysis The bitches were randomly allocated into the test and control groups using the lottery method. Fisher’s exact test and the ‘‘t’’ test were applied to data as appropriate using a computer program ‘‘STAT-SAK’’ (The Statistician’s Swiss Army Knife, Version 2.40, by Gerard E. Dallal Malden, MA 02148). 2.9. Duration of cervical patency (experiment 1) The passage of radio-opaque medium from the vagina to the uterus was determined each day during simulated estrus until cervical closure. The dogs were anesthetised as described previously. Sixty milliliters of a 1 in 4 solution of ‘Isovue-370’ (Bracco, Regional Health Care Products Group, NSW, Australia) was infused into the genital tract [15] using an intra-vaginal balloon catheter (Lyppard, Cheltenham, Australia) inflated with 50 mL of air. The catheter was left in place for 10 min. A left lateral radiograph was taken before infusion and immediately after infusion, followed by left lateral and ventro-dorsal radiographs 10 min later. These bitches received intra-uterine inoculation with E. coli or with control broth during their subsequent simulated diestrus. 2.10. Intra-vaginal inoculation of E. coli and incidence of CEH/P (experiment 2) One milliliter of the E. coli culture (109–1010colony forming units (cfu)) was inoculated intra-vaginally in each bitch (n = 5) daily throughout the simulated estrus. The hindquarters of each bitch were raised during the inoculation and for 5 min afterwards. The general health and feed intake of each dog was monitored throughout the study. Blood samples for hematology were taken daily prior to the inoculation of the bacteria and on the day of euthanasia. Ultrasonographic examination of the uterus of each bitch was performed every 4 days after the last inoculation. Necropsies were performed 12 days after the last inoculation.
2.11. Intra-uterine inoculation of E. coli and incidence of CEH/P (experiment 3) Ten bitches were allocated into two groups of five. Each group comprised three test and two control dogs. One milliliter of E. coli culture (109–1010cfu) or sterile broth was inoculated (0.5 mL into each uterine horn) on Day 12 (first group) or Day 8 (second group) of simulated diestrus by laparotomy using a 24 gauge intra-venous catheter (Optiva, Johnson & Johnson Jntl, Belgium). The inoculations were performed under the anesthetic regimen described previously. Feed intake, disposition and demeanour were monitored throughout the study in both the groups. In addition, daily water intake was measured after the inoculation in the second group. Blood samples for hematology (collected by jugular venipuncture) were taken prior to inoculation and then every 2 or 3 days after inoculation and on the day of euthanasia. Vaginal smears were taken every 3 days after inoculation (second group) and on the day of euthanasia (both groups). Ultrasonographic examination of uterus was performed on the day of euthanasia (one bitch) and on a day before euthanasia (nine bitches). Necropsies were initially scheduled 14 days after inoculation, or before then if any of the bitches showed any signs of ill health. Necropsies were performed 14 days (first group), 7 days (1/5 bitch, second group) and 9 days (4/5 bitches, second group) after inoculation. 3. Results All the bitches were in anestrus at the commencement of the study as determined by small vulvar size, vaginal cytology (SCI 0%) and plasma progesterone concentrations (<0.5 ng/mL). The vaginal superficial cell index was >50% on Day 7 of the estradiol benzoate treatment and increased to >90% by Day 13 of treatment. Estrus commenced 10–11 days after the start of treatment. 3.1. Duration of cervical patency (experiment 1) Radiographs taken before, immediately after the infusion and 10 min later, demonstrated the entry of radio-opaque fluid into the uterine horns through the cervix. In four of five bitches, the cervix remained open throughout simulated estrus. In one bitch, the cervix closed on Days 3 or 4 of a simulated estrus of 6 days. The degree of patency of the cervix reflected by the amount of fluid in the uterus varied (Table 1, Fig. 1a–c); however, this variation was not related to the stage of estrus.
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
1533
Table 1 The presence of radio-opaque fluid in the uterus 10 min after intra-vaginal infusion in dogs Dog ID
NA6 NA7 NA8 NA9 NA10
Duration of estrus (d) 6 6 5 6 6
Day of estrus 1 + ua u u ++
Day of diestrus 2 ++ + ++ + ++
3 ++ u u u ++
4 + 0 + ++ ++
5 ++ 0 ++ + ++
6 ++ nd ++ + ++
1
4/5 b
nd nd ++ nd nd
0 0 0 0 0
0, nil; +, small amount; ++, large amount. a Unclear. b Not done.
3.2. Intra-vaginal inoculation of bacteria and incidence of CEH/P (experiment 2) All the dogs were alert and bright and showed no change in feed intake throughout the period of
observation. There were no significant differences in leucocyte concentrations in blood samples taken during the period of inoculations and on the day of euthanasia. Ultrasonographic examination of the uterus in each bitch revealed no abnormality. No gross
Fig. 1. (a) Bitch at Day 4 of diestrus. Note that the radio-opaque fluid did not pass the cervix (shown by an arrow) indicating a closed cervix (graded as 0). (b) Bitch at Day 4 of estrus. Note that a small amount of radio-opaque fluid (shown by an arrow) passed through the cervix into the uterine horns (graded as +). (c) Bitch at Day 2 of estrus. Note that a large amount of radio-opaque fluid (shown by an arrow) passed through the cervix into the uterine horns (graded as ++).
1534
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
Fig. 2. (a) The uterus of a bitch 12 days after the daily intra-vaginal inoculation with E. coli during simulated estrus (Group 2). The uterus is small and contains no fluid. There is no evidence of CEH/P. (b) Light micrograph of a transverse section of the uterus of a bitch 12 days after daily intra-vaginal inoculation with E. coli during simulated estrus (Group 2). No signs of inflammatory response or cysts. There is no evidence of CEH/P (40 hematoxylin eosin stain).
pathological changes were observed in the uteri at necropsy (Fig. 2a). The uterine flushings obtained were clear, colourless and watery in consistency. E. coli were not isolated from the uterine flushings. Histologically, the tissue sections showed no signs of CEH/P (Fig. 2b). 3.3. Intra-uterine inoculation of bacteria and incidence of CEH/P (experiment 3) A small reduction in feed intake, slight purulent vaginal discharge and leukocytes in vaginal smears were observed in all the test dogs of first group. The total white blood cell numbers were higher in all the test dogs than in control dogs (P > 0.05) on the day of euthanasia but still remained within the reference range (3.6– 13.3 109 L 1), except one dog where the numbers raised above the reference range (15.6 109 L 1).
Fig. 3. (a) The uterus of a test bitch 9 days after intra-uterine inoculation with E. coli during simulated diestrus (Group 3). Note that both the uterine horns were enlarged and filled with fluid, indicating CEH/P. (b) The uterus of a control bitch 9 days after intra-uterine inoculation with Luria–Bertani broth during simulated diestrus (Group 3). The uterus is small and contains no fluid. There is no evidence of CEH/P.
These animals were alert and bright throughout. In the test dogs of the second group, there were no changes in feed intake, but there was increased water intake. No vaginal discharge and no infiltration of leukocytes in vaginal smears were observed. The total white blood cell numbers were higher in the test dogs than in control dogs (P > 0.05) on the day of euthanasia but remained within the reference range (3.6–13.3 109 L 1). The control dogs of both the groups showed no changes in any of these characteristics. One test dog of the second group became depressed on Day 7 after inoculation. Ultrasound examination revealed a fluid-filled uterus. The dog was euthanized the same day. Other dogs of the second group who remained bright and alert throughout were
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
euthanased the following day. Ultrasonographic examination of the uterus of test dogs of both the groups indicated large, fluid filled uteri. At necropsy, the uteri of test dogs were filled with purulent exudate (measured in second group dogs) of yellow to reddish brown colour and a fetid odour whereas the uteri obtained from the control dogs were grossly normal (Fig. 3a and b). Colonies of E. coli were isolated from the uterine contents of the test dogs and the concentration of E. coli in each milliliter of the purulent exudate was calculated in for each dog (Table 2). The uterine flushings obtained from the control dogs did not produce E. coli colonies on the agar plates. Histological examination of each uterine horn and the body of the uterus revealed large cystic endometrial glands and increased endometrial thickness, an influx of inflammatory cells into the glandular
1535
Table 2 The volume of purulent exudate in the canine uterus and the number of E. coli calculated in each milliliter of purulent exudates Dog ID
T/C a
E. coli colonies (cfub/mL)
Purulent exudate (mL)
NA6 NA7 NA8 NA9 NA10 NA12 NA13 NA14 NA15 NA16
T T T C C T C T C T
1.7 109 1.77 109 2.4 109 Nil Nil 5.6 109 Nil 1.19 1010 Nil 1.49 109
ndc nd nd nd nd 25–26 Nil 15–16 Nil 35
a b c
Test/control. Colony forming units. Not done.
and uterine luminal areas (Fig. 4a) and bacterial colonies. Histological examination of uterine tissues from the control dogs did not reveal any changes in the endometrial glands or any infiltration of inflammatory cells (Fig. 4b). 4. Discussion
Fig. 4. (a) Light micrograph of a transverse section of the uterus of a test bitch 14 days after intra-uterine inoculation with E. coli during simulated diestrus (Group 3). There is severe inflammation with cystic glands (~) containing inflammatory debris and the lumen (*) is filled with inflammatory cells. (40 hematoxylin eosin stain). (b) Light micrograph of a transverse section of the uterus of a control bitch 14 days after intra-uterine inoculation with Luria–Bertani broth during simulated diestrus (Group 3). There is no evidence of CEH/P (40 hematoxylin eosin stain).
The results of this study indicate that intra-uterine inoculation of E. coli during simulated diestrus in ovariectomised bitches can reliably induce CEH/P. This validates the ovariectomised bitch model for CEH/P and was consistent with previous findings of the development of CEH/P resulting from the intra-uterine inoculation of E. coli during diestrus [8,16,17]. The occurrence of CEH/P in the estradiol-progestagen treated ovariectomised bitch model adds further confirmation to the validity of this model for the study of the estrous cycle [10]. The failure of intra-vaginal inoculation to induce CEH/P is similar to previous findings where intravaginal inoculation of E. coli during naturally occurring estrus also failed to induce CEH/P [18]. That study involved only one intra-vaginal inoculation during each estrus; in the present study, daily inoculations of pathogenic E. coli during simulated estrus were performed. The lack of CEH/P in spite of the greater frequency of inoculation with a pathogenic strain of E. coli indicated that the normal, healthy uterus during estrus (simulated or natural) has strong defences and can eliminate large doses of pathogenic E. coli [5]. It also suggests that some other factor(s) may be required to initiate infection and assist the persistence and action of bacteria in the uterus [18].
1536
N. Arora et al. / Theriogenology 66 (2006) 1530–1536
The cervix remained open throughout the simulated estrus in four of five bitches in our model, in contrast to a previous study in ovary-intact bitches where the cervix closed 2.6 1 days before Day 1 diestrus [15]. The methodologies of both the studies involved similar amounts of fluid inoculated into the vagina under similar pressure. The reason for the different finding is unclear, but suggests that bitches in simulated estrus should be more susceptible to uterine infections than bitches in natural estrus because of the increased time and opportunity for bacteria to enter the uterus. The finding of a purulent vaginal discharge in the test dogs of the first group receiving intra-uterine inoculation indicated an open-cervix pyometra, whereas in the second group, the lack of such findings indicated a closed cervix pyometra. This difference may have been due to the fact that the dogs used in first group had previously been used for the part of the study examining the patency of the cervix by performing daily infusions of radio-opaque fluid. This treatment may have affected cervical function. In conclusion, a model for cystic endometrial hyperplasia and pyometra using ovariectomised bitches treated with estradiol benzoate, megestrol acetate and E. coli has been validated. This model can be easily and reliably reproduced and used as a basis for future studies of the pathogenesis and prevention of the disease. Future studies to refine the model will involve reduction of the number of E. coli in the inoculum, use of E. coli without the UVF in early diestrus and intra-uterine inoculations during estrus. Acknowledgments We thank Dr. Ben Landon for his guidance in surgical technique, Dr. Cathy Beck and Sylvia Meekings for their assistance in radiology, Dr. Jenny Charles for histological advice and Fiona Armour, Joanne Allen and Margaret Sung for their technical support. The funding support of the Victorian Canine Association, Canine Research Foundation is gratefully acknowledged. References [1] Dow C. The cystic hyperplasia–pyometra complex in the bitch. Vet Rec 1957;69:1409–15.
[2] Hardy RM, Osborne CA. Canine pyometra: pathophysiology, diagnosis and treatment of uterine and extra-uterine lesions. J Am Anim Hosp Assoc 1974;10:245–68. [3] Kennedy PC, Miller RB. The female genital system. In: Palmer N, editor. Pathology of domestic animals. Academic Press; 1993. p. 349–86. [4] Johnston SD, Kustritz MVR, Olson PNS. Disorders of the canine uterus and uterine tubes (oviducts). In: Kersey R, editor. Canine and feline theriogenology. W.B. Saunders Company; 2001. p. 206–24. [5] Noakes DE, Dhaliwal GK, England GCW. Cystic endometrial hyperplasia/pyometra in dogs: a review of the causes and pathogenesis. J Reprod Fertil Suppl 2001;57:395–406. [6] Nomura K. Induction of a deciduoma in the dog. J Vet Med Sci 1994;56:365–9. [7] Bosschere HD, Ducatelle R, Tshamala M. Is mechanically induced cystic endometrial hyperplasia (CEH) a suitable model for study of spontaneously occurring CEH in the uterus of the bitch? Reprod Dom Anim 2002;37:152–7. [8] Bosschere HD, Ducatelle R, Tshamala M. Uterine oestrogen and progesterone receptor expression in experimental pyometra in the bitch. J Comp Pathol 2003;128:99–106. [9] Chen YMM, Wright PJ, Lee CS. A model for the study of cystic endometrial hyperplasia in bitches. J Reprod Fertil Suppl 2001;57:407–14. [10] Chu PY, Lee CS, Moore PF, Wright PJ. Oestrogen and progesterone treated ovariectomised bitches: a model for the study of uterine function. J Reprod Fertil Suppl 2001;57:45–54. [11] Schutte AP. Canine vaginal cytology. III. Compilation and evaluation of cellular indices. J Small Anim Pract 1967;8: 313–7. [12] Holst P, Phemister R. Onset of diestrus in the Beagle bitch: definition and significance. Am J Vet Res 1974;35:401–6. [13] Wright PJ. Practical aspects of the estimation of the time of ovulation and of insemination in the bitch. Aust Vet J 1991;68: 10–3. [14] Chen YMM, Wright PJ, Lee CS, Browning GF. Uropathogenic virulence factors in isolates of Escherichia coli from clinical cases of canine pyometra and feces of healthy bitches. Vet Microbiol 2003;94:57–69. [15] Silva LDM, Onclin K, Verstegen JP. Cervical opening in relation to progesterone and oestradiol during heat in beagle bitches. J Reprod Fertil 1995;104:85–90. [16] Nomura K. Canine pyometra with cystic endometrial hyperplasia experimentally induced by E. coli inoculation. J Vet Sci 1983;45:237–40. [17] Nomura K, Funahashi H. Histological characteristics of canine deciduoma induced by intrauterine inoculation of E. coli suspension. J Vet Med Sci 1999;61:433–8. [18] Nomura K, Yoshida K, Funahashi H, Shimada Y. The possibilities of uterine invasion of Escherichia coli inoculated into the vagina and development of endometritis in the bitches. J Anim Reprod 1988;34:199–203.