Comp Clin Pathol (2006) 14: 186–190 DOI 10.1007/s00580-006-0596-4 ORIGINA L ARTI CLE M. R. Ahmadi . S. Nazifi . H. R. Ghaisari . M. Damchy Evaluation of the cytology of uterus, vagina, and clitoris as predictors of uterine condition in the mare Received: 31 October 2005 / Accepted: 10 January 2006 / Published online: 28 February 2006 # Springer-Verlag London Limited 2006 Abstract Detection of venereal disease and acute or chronic endometritis can be achieved by cytological examination. The presence of significant numbers of neutrophils in the uterine smear is the best indication for equine endometritis. Uterine function may be evaluated by examining clitoral or vaginal smears. In this study, 20 nonpregnant mares were selected and their reproductive functions were evaluated by rectal palpation. The levels of progesterone were measured by radioimmunoassay (RIA). Cytological smears were prepared from clitoris, vagina, and uterus. A wet swab was applied to the clitoral fossa and the collected material transferred to a microscope slide. Vaginal samples were collected using 30 cm wet swabs. Uterine samples were collected from uterine lavage fluid. There was no significant difference (P≥0.05) between cell densities in the estrus phase (p4<1 ng/ml) compared with the interval between estrus phases (p4>1 ng/ml). Significant differences (P<0.05) were noted in epithelial cell, lymphocyte, and neutrophil densities between diestrus phase (p4≥3.5 ng/ml) and other phases (p4<3.5 ng/ml). Epithelial, neutrophil, and lymphocyte densities were significantly different between clitoral, vaginal and uterine smears (P<0.05). There were significant differences in epithelial cells and neutrophils of vaginal and uterine smears in mares affected by metritis compared with repeat breeder and normal mares. We concluded that the clitoral cytology of mares could not predict uterine physiology but evaluation of vaginal cytology showed more appropriate potential. However, further study is needed to assess vaginal cytology in comparison with other methods for evaluating uterine reproductive physiology. Keywords Mare . Endometritis . Cytology . Uterus . Vagina . Clitoris M. R. Ahmadi (*) . S. Nazifi . H. R. Ghaisari . M. Damchy Department of Clinical Sciences, School of Veterinary Medicine, Shiraz University, P.O. Box 1731, Shiraz, Iran e-mail: rahmadi@shirazu.ac.ir Introduction Early studies performed by Knudsen (1964) indicated the potential value of endometritic cytology. Cytological examination of the endometrium is often used in the mare to evaluate causes of infertility, specifically to detect venereal disease and acute or chronic endometritis (Ricketts et al. 1993; Roszel and Freeman 1988). Cytological diagnosis of equine endometritis is based on the presence or absence of significant numbers of neutrophils in the smear. Neutrophils are very rarely seen in normal mares but can often be detected at the “foal heat” or the first estrus of the breeding season in maiden mares although there is no endometritis (Pycock and Newcombe 1996). Also, in many cases, the uterine luminal fluid, which accumulates before mating, is sterile and contains no neutrophils (Pycock and Newcombe 1996). Results of several researchers regarding the presence of neutrophils in uterine fluid are shown in Table 1. Troedsson (1999) suggests that a transient endometritis is a normal physiological response to breeding. However, in mares with impaired uterine defense mechanisms, the condition may develop into a persistent endometritis and subsequent subfertility. Mares with persistent and recurrent endometritis are said to be susceptible to infection; in particular, they are unable to resolve the acute endometritis that always follows mating. It is thought, therefore, that these mares have a local immunological defect in the uterus that impedes the elimination of bacteria. Studies on immunoglobulins, opsonins, and the functional ability of neutrophils in the uterus of susceptible mares have not confirmed the presence of an impaired immune response (Allen and Pycock 1989). For reliable cytological results, a collection technique is required that yields many well-preserved cells representative of a large uterine surface area (Roszel and Freeman 1988). Endometrial cells for cytological evaluation may be collected in uterine aspirations, washings, or by using guarded cotton swabs. A technique using the plastic cap of a guarded culture swab was also described (LeBlanc 1991). 187 Table 1 Quantitative estimation of neutrophils in endometrial smears Authority (Year) Method of quantification Knudsen (1964) Wingfeild Digby (1978) Asbury et al. (1982) Asbury (1984) Couto and Hughes (1984) Brook (1985) Positive if >1 neutrophil in 5 fields (×200) Estimates ratio of neutrophils to epithelial cells Positive if ratio of epithelial cells to neutrophils >10:1 Positive if more than 1 neutrophil in 5 high-powered fields Graded according to ratio of neutrophils to epithelial cells in 8 fields (×100) Negative if no PMNs in 10 fields (×400) Doubtful if 1 to 5 PMNs in 10 fields Positive if >5 PMNs in 10 fields Used a 0 to 5 grading method depending on ratio of PMNs to epithelial cells Positive if >2% of cells were neutrophils Positive if ≥1 neutrophil per field (×400) Negative 0% neutrophils ± less than 0.5% PMNs +0.5 to 5% PMNs 2+5 to 30% PMNs 3+ >30% PMNs La Cour and Sprinkle (1985) Ball et al. (1988) Purswell et al. (1989) Rickets and Mackintosh (1989) PMN Polymorphonuclear The activity of uterine-derived cells of susceptible mares was unaffected by stage of cycle. Susceptibility to endometritis was not associated with a defect in the phagocytic function of uterine neutrophils. Also, the function of uterine cells from resistant mares was greater during estrus than diestrus (Asbury and Hansen 1987). However, there are no reports regarding cytological evaluation of clitoris and vagina in comparison with uterine cytology for mares. The aim of this study was (1) to compare the density of reactive cells in clitoral, vaginal, and uterine smears of mares; (2) to detect differences in cells densities between different progesterone levels; and (3) to evaluate the use of clitoral or vaginal smears for the detection of uterine status. Materials and methods Twenty nonpregnant mares were selected randomly from riding stud horses around Shiraz. Reproductive functions were evaluated by rectal palpation in middle breeding season and mares had not been mated in the preceding 30 days. Cytological smears were prepared from clitoris, vagina, and uterus. The tail was bandaged and held to one side and the clitoris sample was collected by bacteriological swab. The wet swab was applied to the clitoral fossa and the collected was material transferred to a microscope slide. The vulva and perineum were washed thoroughly with clean water. The area was dried using a disposable paper towel. The vaginal sample was collected using a 30cm wet swab. The wet swab was applied to the vagina and the collected material transferred to a microscope slide. The uterine sample was collected, using sterile technique, from uterine lavage fluid. The uterus fluid was collected using an equine uterus catheter (Equine Uterus Catheter, Sopa, Iran). The balloon catheter was placed through the cervix, which was inflated with 60 ml of air to ensure retention within the uterus. Normal saline (100–200 ml at 40°C) was injected and the uterus was manipulated by rectal palpation to distribute the fluid into both uterine horns. As much of the saline as possible was then recovered and collected into a 250-ml cylinder for 30–40 min, after which the supernatant was removed to leave approximately 20 ml. This was centrifuged at 250×g for 5 min. The resultant sediment suspension was used to prepare smears to be stained with Diff-Quik for cytologic examination. The smears were labeled, air-dried, fixed with methyl alcohol at five min, and stained with Giemsa for 20 min followed by washing in distilled water and subsequent drying. At least 100 to 200 cells were evaluated by light microscope and were counted over 20 microscopical fields (×900). Differential cellular counts were carried out on Giemsa-stained smears of the mucosa. The counted cells were epithelial, neutrophil, lymphocyte, macrophage, and monocyte cells. Blood samples were collected from the jugular vein and serum was separated by centrifugation at 750×g for 15 min. Sera were frozen and stored at −20°C. Progesterone level was measured by RIA (Orion Diagnostica, Finland) with sensitivity of 0.2 ng/ml and the intra- and interassay were 10.2 and 6.5%, respectively. The mares were divided to two groups by progesterone levels. First, mares were separated by progesterone level below 1 ng/ml (estrus phase) and above 1 ng/ml. Second, the mares were divided to bellow 3 ng/ml and above 3.5 ng/ml (diestrus phase) (Ginther 1993). Mares were also grouped as repeat breeder mares (infertile for at least one breeding season) and mares who showed signs of endometritis on clinical examination. Data were analyzed by SPSS software, version 11.5. The independent sample t test was used to detect significant differences between clitoral, vaginal, and uterine cell densities in mares with different progesterone levels. Cell densities of clitoris, vagina, and uterus in normal, clinical endometritis and repeat breeder mares were compared using one-way ANOVA and Duncan’s multiple range tests. 188 Results There was no significant difference (P≥0.05) in cell densities of mares at estrus phase (p4<1 ng/ml) compared to mares at other phases (p4>1 ng/ml) (Table 2). There was significant difference (P<0.05) in epithelial cell, lymphocyte, and neutrophil densities when mares were divided to midluteal phase (p4>3.5 ng/ml) and other phases (p4<3.5 ng/ml). The neutrophil densities in vaginal and uterine smears were significantly more than in clitoral smears (P<0.05). Also, epithelial cells and lymphocytes in clitoral and vaginal smears were significantly increased compared to uterine smears (P<0.05) (Table 3). Neutrophils in vaginal smears and uterine lavage smears of mares affected by endometritis were significantly higher (P<0.05) than repeat breeder and normal mares but epithelial cells in repeat breeder mares were significantly lower than in other groups (P<0.05). There was significant difference in monocytes of clitoral smears between the three groups (P<0.05), but in clitoral smears, there was no significant difference in other cells between the groups (P>0.05) (Table 4). Discussion The results of this study did not show significant differences in cellular densities in smears prepared from clitoris, vagina, and uterus between mares divided by being above or below 1 ng/ml of progesterone. When mares were divided to above and below 3.5 ng/ml of progesterone, there were differences in the densities of epithelial cells and lymphocytes in smears from clitoris and vagina. Also, there was significant difference in neutrophil densities in vaginal and uterine smears when the mares were divided to diestrus (p4>3.5 ng/ml) and other phases. These finding are in agreement with reports by Asbury and Hansen (1987). Significant differences were found in the means of neutrophils in cervical mucosa smears at different estrus cycle phases of ten Holstein dairy heifers. The means (±SD) of neutrophil densities were 0.10±0.10, 17.60±8.83, 0.50±0.30, and 3.60±1.44 at estrus, metestrus, diestrus, and proestrus, respectively (Ahmadi et al. 2000). Ahmadi et al. (2004) found no significant difference in cell densities between cervical mucosa and uterine swab in cattle. The ability of uterine fluid samples to attract neutrophils was measured using chemotactic chambers. There was no significant difference between distances migrated by neutrophils toward fluids obtained during estrus or diestrus. Chemotaxis scores tended to be higher with samples from infected uteri and the difference was significant for samples from two mares. Chemotaxis was neither significantly correlated with protein concentration of uterine fluid nor with serum estrogens or progesterone (Blue et al. 1984). Saltiel et al. (1987) considered cervico-endometrial cytology in mares. The number of neutrophils was positively correlated with the percentage of eosinophils, bacterial flora, and cellular necrosis but had a negative association with the presence of ciliocytopholia. The significantly higher percentages of monocytes in clitoral smears in mares affected by metritis in comparison with repeat breeder mares revealed that the presence of leukocytes in the clitoris may originate from uterine infections if the mare has discharge. The significantly higher neutrophil densities in mares affected by clinical endometritis shown in vaginal swabs and uterine lavage smears compared with normal and repeat breeder mares showed a role for cytological evaluation of vagina and uterus. The mean percentage of neutrophils in normal and repeat breeder mares were lower than 5% in vaginal and uterine smears, but were higher than 30% in mares affected by endometritis. Significantly higher neutrophil densities in vaginal smear and uterine lavage smears of mares affected by endometritis revealed that neutrophil-positive vaginal cytology can be useful in reproductive tract evaluation. This was supported by the lack of significant difference between cytology evaluation of vagina and uterus. However, cytological evaluation of vagina and uterus is not usually useful in repeat breeder mares because repeat breeder syndrome in mares is due to many factors and Table 2 Comparison of mean (±SD) cytology of clitoris, vagina, and uterus of mares with different progesterone (p4) levels Cells Groups Number Clitoris Vagina Uterus Epithelial A B A B A B A B A B A B 10 3 10 3 10 3 10 3 10 3 10 3 96.50±2.22 95.67±0.58 0.00±0.00 0.33±0.58 0.40±0.70 0.33±0.58 0.10±0.31 0.33±0.58 2.40±1.07 3.00±1.00 0.60±0.84 0.33±0.58 96.30±2.31 95.33±1.53 0.40±0.70 1.00±1.00 0.30±0.67 0.00±0.00 0.10±0.32 0.67±1.15 2.30±1.25 2.33±1.15 0.70±0.82 0.67±0.58 96.90±1.79 97.33±1.15 0.40±0.70 0.67±1.15 0.20±0.42 0.00±0.00 0.10±0.32 0.33±0.58 1.80±1.03 1.33±1.15 0.60±0.97 0.33±0.58 Neutrophil Basophil Eosinophil Lymphocyte Monocyte A=P4>1 ng/ml and B=P4<1 ng/ml 189 Table 3 Comparison of mean (±SD) cytology of clitoris, vagina, and uterus of mares with different progesterone (p4) levels Cells Groups Number Clitoris Vagina Uterus Epithelial C D C D C D C D C D C D 5 8 5 8 5 8 5 8 5 8 5 8 97.80±1.30a 95.38±1.77b 0.00±0.00a 0.13±0.35a 0.20±0.45a 0.50±0.76a 0.00±0.00a 0.25±0.46a 1.80±0.84a 3.00±0.93b 0.20±0.45a 0.75±0.89a 97.60±1.52a 95.13±1.96b 0.00±0.00a 0.88±0.83b 0.60±0.89a 0.00±0.00a 0.00±0.00a 0.38±0.74a 1.40±0.89a 2.88±0.99b 0.40±0.55a 0.88±0.83a 97.60±1.14a 96.63±1.85a 0.00±0.00a 0.75±0.89b 0.20±0.45a 0.13±0.35a 0.00±0.00a 0.25±0.46a 1.80±0.84a 1.63±1.19a 0.40±0.55a 0.63±1.06a Neutrophil Basophil Eosinophil Lymphocyte Monocyte Values of each cell with different superscripts in each column are those which differ significantly (P<0.05) C=P4>3.5 ng/ml and D=P4<3 ng/ml endometritis in one of several factors that can cause infertility in mares. These results are in agreement with reports of Ricketts and Mackintosh (1989). Results of uterine culture are best interpreted in relation to numbers of neutrophils detected at the endometrial surface by exfoliative cytology. However, mares with chronic endometritis may have little surface irritation (Roszel and Freeman 1988). If neutrophils are present but there is an absence of pathogen growth on the culture, this most likely indicates the presence of a pathogen that was missed during sampling (Ball et al. 1988). Ball et al. (1988) considered a low volume (60-ml) uterine flush using phosphate buffered saline via a transcervical balloon catheter in barren or subfertile mares. In practice, this more complicated sampling method (compared with the guarded swab) would only be appropriate for those mares that were previously proven to be subfertile or where a previous sample using a guarded swab has resulted in absence of a pathogen in culture but the presence of neutrophils in the cytology smear. Collection of a surface sample of the endometrial lining of the uterus with the guarded endometrial swab is undoubtedly the easiest method of sampling. Bourke et al. (1997) compared cytological findings from cotton swab and cytology brush to histopathology findings. The cytology brush technique provided samples with higher cell yields and fewer distorted epithelial cells, providing a more satisfactory sample for cytological evaluation. Although polymorphonuclear (PMN) numbers were not significantly different by either technique, more studies are Table 4 Cytologic comparison of clitoris, vagina, and uterus of normal mares, clinical endometritis, and repeat breeder mares (mean±SD) Cells Groups Number Clitoris Vagina Uterus Epithelial N E R N E R N E R N E R N E R N E R 13 4 3 13 4 3 13 4 3 13 4 3 13 4 3 13 4 3 96.31±1.97 92.25±7.54 96.67±3.06 0.28±0.08 3.00±1.00 0.00±0.00 0.38±0.65 0.00±0.00 0.00±0.00 0.15±0.38 0.75±0.96 0.33±0.58 2.54±1.05 2.00±0.00 2.00±2.00 0.54±0.78a 2.00±1.41b 0.67±0.58a 96.08±2.14a 57.00±8.16b 93.33±5.69a 0.36±0.78a 36.25±4.19b 1.67±2.08a 0.23±0.60 0.00±0.00 0.00±0.00 0.23±0.60 0.96±0.75 0.33±0.33 2.54±1.05 5.00±4.50 2.00±2.00 0.54±0.78 2.00±1.41 0.67±0.58 97.00±1.63a 47.50±5.45b 72.92±5.69a 0.15±0.38a 47.25±9.46b 2.00±3.46a 0.15±0.38 0.25±0.50 0.00±0.00 0.15±0.38 0.25±0.50 0.33±0.58 1.69±1.03 4.75±4.86 2.00±1.73 0.54±0.88 0.00±0.00 0.33±0.58 Neutrophil Basophil Eosinophil Lymphocyte Monocyte Values of each cell with different superscripts in each column are those which differ significantly (P<0.05) N Normal Mares, E mares affected by endometritis, and R repeat breeder mares 190 needed to establish criteria for the interpretation of cytobrush samples in the mare. Endometrial cytological examination is a rapid reliable aid for diagnosing acute endometritis in the mare (Reiswig et al. 1993). Reiswig et al. (1993) reported the correlation between cytology and culture comparable to other reports. Cytology was determined to be as acceptable as biopsy in the determination of acute inflammatory responses associated with bacterial endometritis; however, it was not possible to detect the chronic inflammatory or fibrotic changes by cytology that were revealed by histopathology. Results of this study show that the combination of cytological evaluation and progesterone level detection is probably more useful than cytological evaluation alone. The presence and density of neutrophils in clitoral, vaginal, and uterine flush smears at all phases of normal mares was consistent with almost all previous publications regarding cytological evaluation of equine uteri. We concluded that clitoral cytology could not predict uterine physiology but evaluation of vaginal cytology could add significantly to diagnostic robustness. It should also be noted that false negatives can occur in the evaluation of equine uterine cytology (Wingfield Digby 1978). Further study is needed to explore the potential for vaginal cytology as a predictor for equine uterine conditions. References Ahmadi MR, Nazifi S, Ghaisari HR (2000) Cytology changes in heifer’s cervical mucosa at different phases of the oestrus cycle. 14th ICAR, Stockholm Ahmadi MR, Nazifi S, Khodakaram Tafti A (2004) Cytological studies of uterus and uterine cervical mucosa by swab and aspiration methods in cows. J Fac Vet Med Univ Tehran 59:313–318 Allen WE, Pycock JF (1989) Current views on the pathogenesis of bacterial endometritis in mares. Vet Rec 125:298–301 Asbury AC (1984) Endometritis diagnosis in the mare. Equine Vet Data 5:166 Asbury AC, Hansen PJ (1987) Effects of susceptibility of mares to endometritis and stage of cycle on phagocytic activity of uterine-derived neutrophils. J Reprod Fertil Suppl 35:311–316 Asbury AC, Schultz KT, Klesius PH, Foster GW, Washburn SM (1982) Factors affecting phagocytosis of bacteria by neutrophils in the mare’s uterus. J Reprod Fertil Suppl 32:151–159 Ball BA, Shin SJ, Patten VH, Lein DH, Woods GL (1988) Use of low-volume flush for microbiologic and cytologic examination of the mare’s endometrium. Theriogenology 29:1269–1283 Blue HB, Blue MG, Kenney RM, Merritt TL (1984) Chemotactic properties and protein of equine uterine fluid. Am J Vet Res 45:1205–1208 Bourke M, Milis JN, Barnes AL (1997) High power fields segmented neutrophils. Aust Vet J 75:755–758 Brook D (1985) Cytological and bacteriological examination of the mare’s endometrium. Equine Vet Sci 5:16–22 Couto MS, Hughes JP (1984) Technique and interpretation of cervical and endometrial cytology in the mare. Equine Vet Sci 4:265–273 Ginther OJ (1993) Reproductive biology of the mare, 2nd edn. Equiservices, Wisconsin, pp 232–290 Knudsen O (1964) Endometrial cytology as a diagnostic aid in mares. Cornell Vet 54:415–422 LeBlanc MM (1991) Diseases of the reproductive system: the mare. In: Colahan PT, Mayhew IG, Merritt AM, Moore JN (eds) Equine medicine and surgery, 4th edn. American Veterinary, Goleta, pp 956–963 La Cour A, Sprinkle TA (1985) Relationship of endometrial cytology and fertility in the broodmare. Equine Pract 7:28–36 Purswell BJ, Ley WB, Sriranganathan N, Bowen JM (1989) Aerobic and anaerobic bacterial flora in the postpartum mare. Equine Vet Sci 9:141–144 Pycock JF, Newcombe JR (1996) The relationship between intraluminal uterine fluid, endometritis and pregnancy rate in the mare. Equine Pract 18:19–22 Reiswig JD, Threlfall WR, Rosol TJ (1993) A comparison of endometrial biopsy, culture and cytology during oestrus and dioestrus in the horse. Equine Vet J 25:240–241 Ricketts SW, Mackintosh ME (1989) Uterine cytology. In: Brook D (ed) [McKinnon AO, Voss JL (eds) (1993)] Equine Reproduction. Lea and Febiger, London, pp, 246–254 Ricketts SW, Young A, Medici EB (1993) Uterine and clitoral cultures. In: McKinnon AO, Voss JL (eds) Equine reproduction. Lea and Febiger, London, pp, 234–245 Roszel JF, Freeman KP (1988) Equine endometrial cytology. Vet Clin North Am Equine Pract 4:247–262 Saltiel A, Gutierrez A, de Buen-Llado N, Sosa C (1987) Cervicoendometrial cytology and physiological aspects of the postpartum mare. J Reprod Fertil Suppl 35:305–309 Troedsson MH (1999) Uterine clearance and resistance to persistent endometritis in the mare. Theriogenology 52:461–471 Wingfield Digby NJ (1978) Studies of endometrial cytology in mares. Equine Vet J 10:167–170