Prevalence, Risk Factors and Impact of Subclinical Endometritis on Reproductive Performance of Nili-Ravi Buffalo

Subclinical endometritis (SCE) is the inflammation of endometrium without systemic illness, hence remains mostly undiagnosed and untreated. The early diagnosis necessitates the evaluation of important risk factors. The objective of this study was to identify the risk factors of SCE and their effects on reproductive performance of Nili-Ravi buffalo. 100 buffaloes approaching parturition were selected. During calving, various risk factors viz., type of calving, peri-parturient disorders, sex and birth weight of calf, gestation period and season of calving were recorded. Buffaloes were subjected to endometrial cytology using cytobrush method on 45 day postpartum (DPP) and divided into 2 groups viz., buffaloes ‘with SCE’ (>5% PMN; n=38) and ‘without SCE’ (≤5% PMN; n=62). Buffaloes in estrus were artificially inseminated and fertility parameters were recorded. The occurrence of SCE was significantly affected by calving assistance (OR=11.74; P<0.001), peri-parturient disorders (OR=6.87; P<0.05) and gestation period (OR=1.16; P<0.05). Sex, birth weight of calf and season of calving were not associated with SCE. The service period of buffaloes did not vary between two groups. The median days open was significantly (P<0.05) higher in buffaloes with as compared to buffaloes without SCE (141 vs. 117 d). Buffaloes with SCE had significantly (P<0.05) lower first service conception (21.1 vs. 43.5%) and took more (P<0.05) mean number of services per conception (2.71 vs. 1.62) compared to buffaloes without the SCE. It may be concluded that risk factors around the time of calving control the onset of SCE which in turn has negative impact on reproductive performance of Nili-Ravi buffaloes.

Subclinical endometritis (SCE) is the inflammation of the endometrium without systemic signs of illness or clinical infection (Barlund et al., 2008). Hence, the disease condition remains mostly undiagnosed and untreated. As compared to cattle, buffaloes have higher incidence of uterine infections including subclinical endometritis. Ptaszynzka (2003) attributed the reasons for higher incidence of uterine infections in the buffaloes to poor hygiene, morphometry of vulval lips, vaginal stimulation for milk let down and wallowing habit. Diagnosis of subclinical endometritis in bovine species has been hampered by lack of a universally accepted definition of disease and, simple and effective diagnostic techniques. Subclinical endometritis has also been described as 'cytological endometritis' due to an increased proportion of ploymorphonuclear (PMN) cells in endometrial cytology samples obtained by endometrial cytobrush or low-volume uterine lavage (Dubuc et al., 2010). In cows and buffalo, several tests have been developed to evaluate the inflammatory processes of the endometrium viz., histopathology (Pascottini et al., 2016), endometrial cytology (Kasimanickam et al. 2004;Singh et al., 2018;Nehru et al., 2019), cytology of small volume uterine lavage (Gilbert et al., 2005;Dar et al., 2015) and optical density of uterine lavage (Machado et al., 2012).
Cytological assessment using cytobrush technique is the reference method for diagnosis of subclinical endometritis, owing to ease of collection of samples, quality of the samples, rapid results, and repeatability of the test (Wagener et al., 2017). The vast variety of diagnostic cut-off values (5 to 25% PMN cells), to differentiate affected and unaffected animals taking into account different days of sampling after calving (Kasimanickam et al., 2004;Barlund et al., 2008;Galvao et al., 2009), is the only limitation of cytobrush based evaluation. The wide range of PMN thresholds suggest that several animal-level and herd-level risk factors exist that affect the prevalence of this disease. Since the disease condition mostly remains undiagnosed. Identifying risk factors for SCE may allow for dairy management interventions to aid in controlling this costly disease. Over the last decade, consensus has been arrived vis-a-vis proportion of PMN cells in endometrial cytosmears. Accordingly, presence of >5% PMN at 40-60 days postpartum (DPP) is universally accepted as diagnostic threshold for SCE in bovine species (Gilbert et al., 2005).
In recent years, several researchers have reported a negative impact of SCE in dairy cattle and buffaloes on subsequent reproductive performance. Cows diagnosed with SCE had prolonged days open (Kasimanickam et al., 2004;Gilbert et al., 2005;Barlund et al., 2008), and a negative correlation between first service conception rate in cows and PMN count was also reported (Kaufmann et al., 2009). A recent study in buffaloes found drastic reduction in first service conception rate, increase in number of services per conception and increase in time to pregnancy in buffaloes with SCE (Nehru et al., 2019). The extra days without conception (Kasimanickam et al., 2004;Gilbert et al., 2005;Barlund et al., 2008), extra artificial inseminations (Gilbert et al., 2005;Barlund et al., 2008), and subsequent culling result in huge economic losses.
Our objectives were to (1) determine risk factors for SCE, and (2) evaluate the effects of SCE on reproductive performance in an organized herd of Nili-Ravi buffalo.

Experimental animals
The present study was conducted at the Central Institute for Research on Buffaloes, Nabha, in the Punjab State of India. The experimental station is located at 30°22´ N latitude and 76°12´ E longitude, having tropical climate characteristics where the weather is hot and sub-humid for most of the year, yearly temperature ranges from a minimum of 2°C to a maximum of 48°C and yearly relative humidity ranges from a minimum of 50% to a maximum of 85%.
Buffaloes in 2 nd -5 th parity approaching calving were identified and kept under close observation until the day of calving. During calving, various risk factors such as type of calving (normal or dystocia), peri-parturient disorders (cervico-vaginal prolapse, uterine prolapse, retained placenta, etc), sex of calf, body weight of calf, gestation period and season of calving were recorded. Seasons were categorized as follows: summer -March to June; rainy -July to October; and winter -November to February. All the selected buffaloes were closely monitored and were subjected to endometrial cytobrush based cytology on 45 DPP.

Endometrial cytology
Pap smear brush for human use (Mayfair Surgical Corporation, Ludhiana) fixed on the tip of stylette and passed through a bovine uterine swab catheter was used as cytobrush assembly. The cytobrush assembly was passed through cervix and guided past the uterine bifurcation into previously gravid uterine horn (Barlund et al., 2008). The pap smears was pushed out through the anterior end of uterine catheter and rotated across the endometrium to collect cellular material. The cytobrush was withdrawn and smears were prepared. The cytosmears were fixed with methanol and stained with Modified Wright Giemsa Stain (Sigma-Aldrich Inc., USA) as per the guidelines. Slides were examined using light microscopy under 400 × magnification to identify endometrial epithelial and PMN cells. A total of 300 cells per slide were counted (Melcher et al., 2014).

Allocation into groups
The threshold value for percentage of PMN cells in endometrial cytology for diagnosis of subclinical endometritis was adopted as reported by Gilbert et al. (2005). Accordingly at 45 DPP, buffaloes with >5% PMN cell count in endometrial cytosmears were designated as positive (n = 38) and buffaloes with ≤5% PMN cell counts were designated as negative (n = 62) for subclinical endometritis (Fig. 1). Buffaloes of both the groups were observed for signs of first postpartum overt estrus. Buffaloes in estrus were artificially inseminated (twice as per AM/PM rule) and fertility parameters (up to 120 DPP) were recorded. Pregnancy was confirmed by ultrasonography at day 45 post-insemination.

STATISTICAL ANALYSIS
The statistical analysis was carried out using SPSS Statistics 16.0 (SPSS, Chicago, USA). For logistic regression model, subclinical endometritis (without = no, with = yes), periparturient disorders (0 = no, 1 = yes), type of calving (0 = normal, 1 = dystocia), sex of calf (0 = female, 1 = male), season of calving (0 = summer, 1 = summer, 2 = rainy), in addition to variables like body weight of calf (kg) and gestation period (days) were included as factors. Fertility parameters, viz. calving to first estrus and calving to conception, were analyzed using Kaplan-Meier survival curve analysis, and median values were compared using log-rank test. Buffaloes were censored if not diagnosed as being pregnant before culling, death, or the end of the data collection period, which was 200 DPP. For logistic regression and survival analyses, CI was set at 95%. For all statistical analyses level of significance was set at α = 0.05.

RESULTS AND DISCUSSION
The overall prevalence of SCE was 38% (38 affected and 62 unaffected). Buffaloes with SCE (>5% PMN) had significantly (P<0.001) higher polymorphonuclear cell count (PMN %) at 45 DPP as compared to buffaloes without SCE (≤5% PMN) (Fig. 2). Based on endometrial cytology (>5% PMN), the incidence of subclinical endometritis in Indian buffaloes has been reported to be between 23 (Gahlot et al., 2017) and 33.7% (Nehru et al., 2019); however, a slightly higher incidence based on uterine lavage cytology has also been recorded (41.7%) (Dar et al., 2015).  According to our findings, abnormal parturition, periparturient disorders (cervico-vaginal prolapse, uterine prolapse, retained placenta, etc) and increased gestation period were associated with an increase in prevalence of SCE at 45 DPP ( The main predisposing factors related to the incidence of uterine infections were calving assistance, twin births, malpresented calves and retained placenta (Bell and Roberts, 2007). Similarly, Kasimanickam et al. (2004) reported that cows with peripartum reproductive events (retained placenta, twins, assisted calving) were 3.15 times more likely to have SCE at 20-33 DPP and 3.18 times more likely to have SCE at 34-47 DPP. Peri-partum obstetrical conditions were associated with delay in uterine involution, expulsion of lochia, closure of cervix and regeneration of endometrium (Sheldon et al., 2008). It has been observed that dystocia and retained placenta lead to decrease in phagocytic activities of uterine and peripheral blood neutrophils (Paisley et al., 1986;Lewis, 1997). This may result in increasing the decreased antimicrobial activity and persistence of uterine infection.
Although, sex of calf (male calf) was significantly associated with occurrence of clinical endometritis (Potter et al., 2010), we found no significant effect. The association of sex of calf with endometrial inflammation in earlier study might be due to increased gestation and higher birth weight of male offspring that increases the predisposition to postpartum obstetrical conditions. Interestingly, we found that increased gestation period was associated with occurrence of SCE, which may be due to overgrown calves that result in dytocia and thus indirectly affecting subclinical endometritis.
Although, season of calving was previously reported to be significantly associated with occurrence of uterine infection (Hossein-Zadeh and Ardalan, 2011), we found no association between season of calving and subclinical endometritis. It was reported that calving occurring between November and April dramatically increased the incidence of clinical infection of the uterus during the first month postpartum. This may be due to the fact that during the rainy and winter seasons, the general health of cows decreases, animal are kept indoors in unhygienic conditions making them more vulnerable to uterine infections (Markusfeld, 1984;Bruun et al., 2002). Study on a large herd (57,301 dairy cows) confirmed the direct relation between calving during winter months and clinical metritis (OR=2.4) (Hossein-Zadeh and Ardalan, 2011). However, our study is in agreement with others which reported no significant influence of calving season on prevalence of clinical endometritis (Kim and Kang, 2003) and subclinical endometritis (Carneiro et al., 2014).
Our second hypothesis was that SCE will have a negative effect on reproductive performance in buffaloes. Although several authors have successfully quantified the effects of SCE on reproductive performance in postpartum dairy cows, this study is one of few attempts to evaluate the impact of SCE on reproductive performance in postpartum buffaloes. Survival analysis showed that the service period (calving-to-estrus interval) of buffaloes with SCE was not In contrast to present study, Dubuc et al. (2012) showed that cows with SCE had delayed resumption of ovarian activity after calving. Similarly, cows with SCE at 42 DPP were less likely to ovulate between 63 to 70 DPP than healthy cows (Burke et al., 2010). Plontzke et al. (2010) also reported no effect of SCE on days to first service. No significant effect of subclinical endometritis on the calving to first service was observed in Murrah buffaloes (Dar et al., 2015;Nehru et al., 2019). We found that buffaloes positive for SCE had significantly (P<0.05) reduced first service conception rate as compared to their negative counterparts ( (Galvao et al., 2009;Dubuc, 2011;Madoz et al., 2013;Vieira-Neto et al., 2014). Our observations were also supported by various authors who concluded that animals diagnosed with SCE had prolonged days open and reduced probability of conception at first AI compared to SCE negative cows (Kasimanickam et al., 2004;Gilbert et al., 2005;Barlund et al., 2008). Subclinical endometritis resulted in reduced proportion of cows pregnant within 200-300 DPP compared to healthy cows (60.0 versus 80.0%, respectively) (Gilbert et al., 2005;Barrio et al., 2015). In contrast, Carneiro et al. (2014) failed to arrive at an association between presence of SCE (32-70 dpp) and first service conception rate and pregnancy rate at 150 days postpartum in cows in Brazil.
The exact mechanism by which SCE decreased conception per AI and increased pregnancy loss was not clear. There was mounting evidence that cows with SCE had altered embryo quality and endometrial function Hailemariam et al., 2014). Inflammation in the endometrium had been shown to reduce fertilization in single ovulating postpartum dairy cows (Cerri et al., 2009). Soto et al. (2003) suggested that mediators of the inflammatory cascade, including cytokines impaired early embryonic development and might be part of the mechanism by which fertility is depressed in cows suffering from inflammatory diseases. At molecular level, cows diagnosed with SCE have altered endometrial and embryonic gene expression . Whether SCE per se was the causative agent of changes in endometrial and embryonic gene expression or the buffaloes that developed SCE had underlying factors that also cause changes in the transcriptome remains to be elucidated.
We conclude that the key drivers that determine outcome of subclinical endometritis are as follows: type of calving, peri-parturient obstetrical conditions and gestation period. It is clear that these risk factors around the time of calving control the onset of subclinical endometritis which in turn has drastic impact on reproductive performance of Nili-Ravi buffaloes.