Assessment of the bovine uterus with endometritis using Doppler ultrasound

Postpartum uterine diseases such as metritis and endometritis are highly prevalent in dairy cows. These diseases negatively affect the reproductive performance and consequently economic activity. Cows in the puerperal period presenting endometritis may have alterations in the hemodynamics of the uterine tissue and the uterine arteries, which differ them from healthy cows. Therefore, this study aimed to use the Doppler ultrasonography to describe the hemodynamic changes in the uterus of cows showing endometritis diagnosed between 25 - 35 days postpartum. Eighty-nine Holstein Friesian females with 25 to 35 days postpartum were studied. Cows were assigned to two experimental groups, infected or not infected, according to the results of the endometrial cytology. Clinical examination, vaginoscopy, Doppler ultrasound and sample collection (saline solution was injected and recovered by endoscopie method aiming cytological and microbiological evaluation of the uterus) were also performed. Cows with endometritis had the cervix (P = 0.040) and the left horn (P = 0.020) increased compared to healthy cows. 78.6% of the endometritic cows showed abnormal uterine discharge, while 57.6% of healthy cows had this same condition (P = 0.0005). The spectral Doppler evaluation of the uterine arteries revealed no differences between groups. Bacillus spp., Trueperella pyogenes, Escherichia coli and Staphylococcus intermedius were the most isolated bacteria among samples. Higher score or increase of uterus vascularization of the endometrial Doppler was correlated with Trueperella pyogenes (P = 0.0003) and intrauterine heterogeneous content (P = 0.0047). Finally, mesometrial Doppler was correlated with endometrial Doppler (P < 0.001), uterine bacteria (P = 0.001) and intrauterine heterogeneous content (P = 0.049). Regarding the evolution of uterine alterations, Doppler ultrasonography provides fast results and is a lesser invasive technique such as uterus biopsie and endometrial citology and gives answers about fertility and uterus health.


104
After the conventional ultrasonographic evaluation, hemodynamic patterns of the left and 105 right uterine arteries were analyzed using the spectral Doppler mode [12]. Maximum flow velocity, 106 pulsatility index and systole/diastole ratio were determined as described by Heppelmann, Krüger and 107 Leidl (2013) [17] (Figure 3).

108
Color Doppler mode was used to evaluate the hemodynamic pattern of the uterine tissue in

114
The ovaries were evaluated for their size, consistency and structures (CL, follicles and cysts).
115 Morphological characteristics of the ovaries were assessed to verify the presence of abnormalities 116 such as cysts or tumors, to determine the existence of cyclic luteal ovarian activity and to estimate 117 the probable phase of the cycle, factors that may affect animal's fertility [20]. Particularly, the 118 vascularization of the CL was assessed by colorimetric Doppler mode, and it was classified as 119 containing or not containing vascularization ( Figure 9).  Five hundred microliters of the recovered saline were added to thioglycolate broth (Difco®) 223 and kept at 2°-10 °C until processing at the Laboratory of General Bacteriology, Biological Institute, 224 São Paulo. Ten microliters were plated in 5% sheep blood agar and incubated for 72 hours at 37 °C 225 under microaerophilic conditions. Morphological characteristics of each colony and Gram staining 226 were performed. Bacteria species were determined using biochemical tests according to Winn et al.

2.8 Statistical analysis
229 Error normality and homogeneity of variance were analyzed by Shapiro-Wilk Test and 230 Bartlett Test, respectively. Non-normal data were analyzed using a non-parametric test (Kruskal-231 Wallis test) (proc npar1way). Differences between the groups (control and endometritis) were 232 evaluated by ANOVA, reaffirmed by a t-test (JMP 12.0 -SAS). Same conditions (proc glimmix from 233 SAS) were used for binary data ("subjective evaluation score of endometrial perfusion" and 234 "subjective evaluation score of mesometrium perfusion"). Correlation between variables were 235 performed using Spearmann correlation test for variables that did not present normal distribution 236 (proc corr from SAS). A significance level of 5% was used for all tests. All tests were performed in 237 SAS.

239 3. Results 240 241
One hundred cows were initially examined, but due to problems with the ultrasound device, 242 eleven samples were lost. So, 89 cows were enrolled in this study, and they were divided accordingly 243 to citology results into two groups: control group (E) (N= 33) and endometritis group (N = 56).

3.1 Ultrasonographic exam 245
Cows with endometritis had the cervix (P = 0.040) and the left horn (P = 0.020) increased 246 compared to healthy cows (Table 1). In the endometritis group, 78.6% of the cows showed abnormal 247 uterine discharge, while in health group only 57.6% of the animals showed this same condition (P = 248 0.0005) ( Table 2). Besides, intrauterine heterogeneous content were increased in sick cows (66.0%) 249 compared to healthy cows (30.3%) (P = 0.0011) ( Table 3). Evaluation of endometrial and 250 mesometrium vascularization by the colorimetric method revealed differences in the vascularization 251 score (vc) of the endometrium between healthy (C) and endometritis group (C) (P < 0.05) ( Table 4 252 and 5). During ovarian examination, presence or absence of CL were not diferent between group (C) 253 and group (E) ( Table 6). In addition, the absence of vascularizated corpus luteum in the left ovary 254 was increased in the endometritis group (P = 0.029) ( Table 7). The spectral Doppler evaluation of 255 the uterine arteries revealed no differences between groups (Table 8).         297 the most isolated bacteria among samples. Yeast was also detected in 25% of samples (Table 9).  309 and endometritis was also detected (Table 10).   (Table 14). Endometrial Doppler was correlated with Trueperella pyogenes (P 323 = 0.0003) and IUHC (P = 0.0047) (Table 15). Finally, mesometrial Doppler was correlated with 324 endometrial Doppler (P < 0.001), uterine bacteria (P = 0.001) and IUHC (P = 0.049). A negative 325 correlation with the presence of yeast (P = 0.022) was also observed (Table 16).

364
Our results showed differences between the size of the cervix and a larger uterus in cows with 365 endometritis. After delivery, the uterus is greatly enlarged (approximately 8 to 10 kg). In general, the 366 uterine macroscopic involution occurs between 3 to 5 weeks postpartum, when the uterus should 367 weight approximately 0.9 kg and the diameter of the pre-gravid uterine horn should be less than 5 368 cm, total involution of the cervix occurs between 4 to 6 weeks postpartum [25].

386
The presence of vascularization in the endometrium (degrees 1 and 2) was increased in cows 387 with endometritis, suggesting differences between healthy and unhealthy cows. The color Doppler 388 mode provides colored images of the blood fow, allowing estimation of the tissue vascularization 400 Clinical and histologic uterine involution ranges from 21 to 50 postpartum, and this may differs from 401 the recovery of hemodynamic patterns of the uterus that requires more time [30].

402
Advancing the understanding of uterine perfusion, Bollwein et al. (2002) [13] when evaluating 403 pregnant animals, concluded that throughout pregnancy there is a strong increase of blood flow in the 404 uterine arteries. However, they observed that the resistivity index decreased in the first 8 months of 405 gestation and remained at a relatively constant level until the calf was born. It is known that during

409
In healthy cows the uterine blood flow decreased after the complete uterine involution.
410 However, cows with uterine disease presents slow uterine involution, occurring between 45 and 65 411 days postpartum. These results indicate an association between incomplete uterine involution and