{"paper_id":"bf655d17-0b3a-498a-89d6-555e1cbec93c","body_text":"Chronic endometritis (CE) in humans is a condition characterised by asymptomatic inflammation of the endometrium. The normal endometrium has variations in immune cell quantities depending on cyclical changes ( Kitaya  et al.  2016 ). However, in CE, there are alterations in endometrial cytokine production, damage to endometrial function and abnormal patterns of lymphocyte subsets in the endometrium ( Kimura  et al.  2019 ,  Xu  et al.  2020 ). CE is associated with subfertility and reproductive complications such as recurrent implantation failure (RIF) and recurrent pregnancy losses (RPLs) ( Li  et al.  2020 ). A uniform or globally accepted diagnostic criterion for human CE is not available, but it is commonly diagnosed by immunohistochemical staining of CD138 expressing plasma cells within the endometrium ( Puente  et al.  2020 ). While CE presents a potential treatable cause of reproductive failure in humans, its pathophysiology and treatment are still not fully understood ( McQueen  et al.  2014 ).\nConversely, inflammation of the bovine endometrium has been extensively explored due to its well-established negative economic impact on the dairy farming industry and animal welfare ( Lee & Kim 2007 ). Cows can present with several different forms of endometritis. Clinical endometritis is characterised by uterine pus and presents with purulent or mucopurulent uterine discharge accompanied by endometrial inflammation often diagnosed via endometrial cytology ( Sheldon 2020 ). Subclinical endometritis (SCE) is characterised by asymptomatic inflammation of the endometrium, much like human CE, with an increased proportion of polymorphonuclear (PMN) cells, often diagnosed via endometrial cytology ( Sheldon  et al.  2019 ). This subset may also be referred to as cytological endometritis. For the purposes of this review SCE refers to both SCE and cytological endometritis. SCE is the cause of infertility in an estimated 15–67% of dairy cows, and each case causes significant financial burden due to the costs of treatment, impaired reproductive performance and production losses ( Mahnani  et al.  2015 ). Research into dairy cow SCE has highlighted a clear pathophysiology in comparison to human CE, with risk factors, potential infective causes and clinical consequences being better understood than in the human. A diagnosis is carried out via a range of methods, including cytology, vaginoscopy, transrectal palpation, ultrasonography of the reproductive tract and biopsy of the endometrium ( Wagener  et al.  2017 ).\nAnatomically, the human and bovine uteri are structurally similar, consisting of the endometrium, the surrounding myometrium and an external perimetrium of loose connective tissue. The bovine uterus is ‘bicornuate’, with the common uterine body separating into two coiled uterine horns, whereas in humans, the organ is described as ‘simplex’, as it consists of a single common uterine body. At a cellular level, the endometrium in both species consists of a stromal layer that contains glandular tissue, underlying a simple columnar epithelium. In the cow, the endometrium contains caruncles; large, vascularised protrusions that have very little glandular tissue and form the maternal side of the placental connections with the developing embryo ( Simões & Stilwell 2021 ). These are absent in the human. The potential for utilising studies of SCE in the bovine population to further understand human CE has not yet been explored. However, it is increasingly recognised that animal studies provide a wealth of knowledge for translational application in clinical research, with emphasis on the need for systematic reviews of animal studies to provide these new insights ( Hooijmans & Ritskes-Hoitinga 2013 ). This offers an opportunity for the analysis of endometritis through a comparative biology lens using knowledge generated in the dairy cow to inform future pathophysiological studies in humans.\nThe aim of the present study was to systematically review the current knowledge regarding the diagnosis and understanding of the pathophysiology of subclinical endometritis in cows in order to identify diagnostic and pathophysiological pathways that had not yet been explored in human studies. The application of a thematic analysis provided a framework for interpretation of the bovine literature, from the perspective of a human studies, and provided a basis for directing the focus of future studies of CE in the human. Thus, the present study informs on the development of novel strategies for the diagnosis and management of chronic endometritis in humans.\n\nA systematic review with an emergent theme thematic analysis was undertaken. This review was prospectively registered on PROSPERO CRD42022264763.\nAn electronic systematic search of EMBASE, MEDLINE, Scopus and CINAHL from 1990 until November 2021 was performed of the literature on SCE or cytological endometritis in the cow. The search strategy included MESH headings; endometritis, reproductive health, female infertility, cow, bovine, reproduct*, chronic endometritis and cow*. No limits or filters were used. The search strategy and number of search results are presented in  Supplementary material  1 (see section on supplementary materials given at the end of this article). In addition, papers provided by expert authors were also included to further contextualise thematic discussion (EW and OBP).\nEligibility assessment of studies by title and abstract was performed independently by two reviewers (KK and JO), with discrepancies resolved by discussion. We included studies that reported on the pathophysiology, diagnostic techniques and associated factors of SCE in dairy cows. Any studies of non-dairy cows,  ex vivo  or  in vitro  studies, symptomatic endometritis cases, case reports and reviews were excluded. There were no restrictions on farm type, setting or location. Full-text review to confirm eligibility was performed by KK and JO. Data extraction was performed by KK using a bespoke data extraction tool presented in  Supplementary material  2, with review by a second reviewer where required (JO).\nAn emergent framework thematic analysis was performed to identify key themes within the included literature. Papers were analysed according to the main concepts and aims of each study, and this helped us identify the main emergent themes of the included papers. Each paper was then compared against these emergent themes to allow the identification of subthemes and the systematic building of ideas relating to these subthemes. This methodology was chosen to identify themes from cow studies which could be translated to human theories and studies.\n\nThe literature search identified 3419 studies, of which 342 studies were eligible for full-text review and 44 studies were identified for further analyses. An additional seven texts identified by expert authors were included. Overall, the review comprised 51 studies. An overview of the selection process is shown in  Fig. 1 .\n Figure 1 PRISMA diagram for study selection.\nPRISMA diagram for study selection.\nEmergent theme analysis revealed two key themes: i) diagnostic methods and ii) pathophysiology. While extensively investigated in studies involving cows, these themes remain inadequately and insufficiently explored in human studies.\n\nTwelve papers explored the optimum method for diagnosis of SCE in dairy cattle ( Table 1 ). Thematic analysis identified six key subthemes in the diagnostic literature: i) cytology for diagnosis, ii) cytological assessment compared to histopathology, iii) cytological assessment compared to alternate diagnostic methods of ultrasound or vaginoscopy, iv–vi) the use of reagent test strips, serum markers and bacterial culture ( Fig. 2 ).\n Figure 2 Methods for diagnosis and risk stratification of SCE investigated by the papers included in the study. \n Table 1 Overview of included studies investigating different diagnostic methods for subclinical endometritis in cows. Study Study population Condition(s) investigated Sampling technique Diagnostic method investigated CT US HP CA RS SM BCG Ahmadi  et al.  (2005) 131 slaughtered cows Acute, subacute, chronic EM Cervical swab; cervical aspiration; uterine swab; uterine aspiration; endometrial biopsy ✓ – ✓ – – – – Barlund  et al.  (2008) 221 cows 28–41 days PP CE and SCE Cytobrush; uterine lavage ✓ ✓ – ✓ – – – Pascottini  et al.  (2016 a ) 32 cows pre and post slaughter SCE Cytobrush; endometrial biopsy; uterine lavage ✓ – ✓ – – – – Brodzki  et al.  (2014) 60 cows PP (30 EM, 30 healthy) SCE Cytobrush; uterine lavage ✓ ✓ – – – – ✓ Cheong  et al.  (2012) 563 cows 40–60 days PP SCE Uterine lavage ✓ – – – ✓ – – Egberts  et al.  (2016) 53 cows 64–1018 days PP SCE Cytobrush ✓ – – – – – – Kasimanickam  et al.  (2004) 228 lactating cows SCE Cytobrush ✓ ✓ – – – – – Kaufmann  et al.  (2010) 209 cows up to 35 days PP CE and SCE Serum samples for NEFAs, BHA, bilirubin, urea; cytobrush ✓ – – – – ✓ – Madoz  et al.  (2013) 53 cows 27–56 days PP, and 418 cows 21–62 days PP SCE Cytobrush ✓ – – – – – – Meira  et al.  (2012) 76 cows 21–47 days PP SCE Cytobrush; endometrial biopsy ✓ ✓ ✓ – – – – Nyabinwa  et al.  (2020 a , b ) 466 cows 21–60 days PP CE and SCE Cytotape ✓ – – ✓ – – – Ricci  et al.  (2017) 70 healthy cows 30–60 days PP SCE Uterine lavage; serum sample for haptoglobin ✓ – – – ✓ ✓ ✓ BCG, bacterial culture or genotyping; CA, clinical assessment (e.g. external inspection, vaginoscopy); CE, clinical endometritis; CT, cytology; EM, endometritis; HP, histopathology; PP, post partum; RS, reagent strips; SCE, subclinical endometritis; SM, serum metabolites; US, ultrasound.\nMethods for diagnosis and risk stratification of SCE investigated by the papers included in the study.\nOverview of included studies investigating different diagnostic methods for subclinical endometritis in cows.\nBCG, bacterial culture or genotyping; CA, clinical assessment (e.g. external inspection, vaginoscopy); CE, clinical endometritis; CT, cytology; EM, endometritis; HP, histopathology; PP, post partum; RS, reagent strips; SCE, subclinical endometritis; SM, serum metabolites; US, ultrasound.\nA strong emergent theme was the use of PMN quantification through cytological assessment of the endometrium for diagnosis of SCE in cows ( Fig. 2 ).\nIn practice, this method generally involves passing a cytobrush through the cervix and into the uterus, and by administering gentle pressure per rectum, the operator ensures the brush makes contact with the endometrium and then gently rotates the instrument to ensure a good cell sample is obtained ( Kasimanickam  et al.  2006 ,  Rana  et al.  2020 ).\nAn endometrial cell sample can also be collected for cytological assessment by uterine lavage, or by cytotape – an adaptation of the cytobrush technique where rolled surgical tape collects the endometrial cells instead of a brush ( Pascottini  et al.  2016 b \n ). Sampling can be performed faster and easier via the cytobrush (or cytotape) method compared to uterine lavage ( Brodzki  et al.  2014 ).\nBoth cytobrush and uterine lavage provide samples of similar quality and cellularity, with lavage achieving a 14.3% sensitivity and 84% specificity for the diagnosis of SCE ( Barlund  et al.  2008 ,  Brodzki  et al.  2014 ).\nThe timing and location of cytological sampling are key in ensuring an accurate representation of the endometrium. It has been suggested that cytobrush analysis may require at least two samples from different locations, preferably the uterine horns, to be able to effectively represent the PMN cell levels, although currently a single-site approach remains the standard ( Egberts  et al.  2016 ). There is a risk of a high false-positive rate if the cytological sample is collected too early in the oestrous cycle, due to normal infiltration of PMN cells ( Egberts  et al.  2016 ). Following calving, physiological inflammation of the endometrium occurs in order to facilitate uterine involution and endometrial repair, so the level of PMN infiltration that signifies SCE differs across the postpartum period ( Foley  et al.  2015 ). This is presently overcome by standard practice involving sampling >20 days post partum.\nBiopsy or histopathology was discussed in three studies and was suggested as the definitive diagnostic tool to diagnose SCE with greater sensitivity than cytology ( Barlund  et al.  2008 ). Compared to cytology, histopathology is regarded as a more sensitive diagnostic tool ( Pascottini  et al.  2016 a ) and is therefore considered the most definitive method for identifying tissue inflammation and endometrial lesions ( Meira  et al.  2012 ). However, biopsy is comparatively expensive, time-consuming and has negative implications on subsequent fertility outcomes, which could affect the interpretation of results or effectiveness of treatments tested ( Barlund  et al.  2008 ,  Pascottini  et al.  2016 a ). Thus, using cytological assessments has proven to be less invasive and fairly precise ( Barlund  et al.  2008 ).\nSeven studies used cytological analysis of the endometrium as a baseline for comparison with alternate diagnostic methods. One of the included articles compared a cytobrush technique to ultrasound and vaginoscopy.  Barlund  et al.  (2008)  hypothesised that assessment of intrauterine fluid accumulation may improve the diagnostic capability of ultrasound for SCE. They found that sonographic uterine fluid assessment had poor sensitivity but good specificity for the diagnosis of subclinical endometritis, especially in cases of increasing volumes of intrauterine fluid. Vaginoscopy proved to be rapid and simple but had only moderate agreement with the cytobrush results and required exudate analysis to make a positive diagnosis. This suggests ultrasound and vaginoscopy are more effective techniques for diagnosis of clinical endometritis, where vaginal discharge is prevalent. Thus, by their very nature, these are not appropriate tests for the determination of subclinical endometritis.\nReagent strips are a surrogate test and have the potential for significant benefit in the diagnosis of SCE as they could facilitate an on-the-step ‘point-of-care’ diagnosis, negating the time and financial consequences of laboratory analysis. Their use in this context was explored in two studies, and while there was a positive correlation between the test strips and a diagnosis of endometritis, with leukocyte esterase, pH and protein all increased in cows with SCE diagnosed via cytology, reagent strips were considered a poor diagnostic technique due to the low specificity and sensitivity seen across both studies ( Cheong  et al.  2012 ,  Ricci  et al.  2017 ).\nSerum markers to identify risk of SCE were explored in two studies. Urea and non-esterified fatty acids (NEFAs) were investigated as markers of risk for SCE. Urea levels were raised at 1 week post partum ( Kaufmann  et al.  2010 ). NEFAs have a sensitivity and specificity of 35% and 89%, respectively, for subclinical endometritis ( Kaufmann  et al.  2010 ), and it has been found that high levels of NEFAs can affect blood PMN functions ( Kaufmann  et al.  2010 ). Serum haptoglobin, a nonspecific acute phase protein that increases in response to inflammation, was examined as a potential marker by one group. However, no significant difference was found in serum haptoglobin between those with positive or negative cytology for SCE ( Ricci  et al.  2017 ).\nA final theme that emerged was bacterial culture obtained from a uterine swab sample. While there is consensus that clinical endometritis in cows is the result of bacterial infection, the association between the presence of specific pathogenic bacteria and SCE is not well established ( Pascottini  et al.  2023 ). Of the included studies, two studies discussed bacterial culture as a diagnostic method for SCE.  Trueperella pyogenes  and  Escherichia coli  were isolated in both studies ( Brodzki  et al.  2014 ,  Ricci  et al.  2017 ); however, a direct association between SCE and the bacterial infection has not been confirmed. The current consensus suggests that SCE is the result of a dysfunctional metabolic and inflammatory response rather than an infectious agent ( Pascottini  et al.  2023 ).\n\nPathophysiology was discussed in 33 papers, and data are presented in  Table 2 . A thematic analysis of the theories surrounding pathophysiology of SCE in cows revealed five key themes: the type of endometritis, metabolic stress, artificial insemination and the postpartum period, and infective causes and cellular pathways. An overview of the identified themes is presented in  Fig. 3 .\n Figure 3 Flow chart of pathophysiology themes and subthemes presented in the papers included in the study. \n Table 2 Overview of included studies investigating the pathophysiology of subclinical endometritis in cows. Study Cows,  n \n Type of EM PMN% Study aims Bacha & Regassa (2010) 59 SCE ≥5% neutrophil count Effect of SCE on reproductive performance Barański  et al.  (2013) 222 SCE 4–18% Impact of EM at week 4 and 6 PP Barrio  et al.  (2015) 94 SCE ≥5% neutrophil count Effect of SCE on reproductive performance Pascottini & LeBlanc (2020) 58 SCE >5% neutrophil count Metabolic markers Pascottini  et al.  (2017) 873 CTE ≥1% PMN Effect of AI on EM status Pascottini  et al.  (2016 a , b ) 512 CTE ≥1% PMN Success of insemination, conception rates with CYTO and without Bruun  et al.  (2002) 2144 Metritis – Risk factors Cheong  et al.  (2011) 779 SCE >10% PMNL Risk factors: herd size, calving pen, postpartum housing, pen moves, parity, ketosis, milk production De Biase  et al.  (2018) 37 CE – \n Coxiella burnetii  detection by PCR Diaz-Lundahl  et al.  (2021) 1648 CTE ≥3.0% PMN Effects of AI on fertility Donofrio  et al.  (2010) – \n § – Activation of  IL8  gene promoter in BoHV-4 infection of bovine ESCs Fábián  et al.  (2008) 131 EM – Bacterial isolates in endometrial samples Fagundes  et al.  (2019) 64 SCE >5% PMN Expression of  CCL5, CXCL8, IL6, IL1B genes Gilbert  et al.  (2005) 141 SCE/CTE 5% neutrophils as cutoff Prevalence and reproductive performance Gobikrushanth  et al.  (2016) 15 CTE >18% PMN ⁋ Reproductive performance, dry matter intake, milk yield in dairy cattle Jiang  et al.  (2021) 3 EM – Describing the role of miR-92B in inflammatory response in bovine EECs, following LPS stimulation. Kasimanickam  et al.  (2006) 275* SCE – Effect of clinical/SCE on cow fertility*** Kudo  et al.  (2021) 14 SCE >1, or a PMN ratio >10% Microbiota of primiparous cows with and without EM Lee  et al.  (2018) 407 CTE 14% at 4 weeks PP Impact on reproductive performance Miller  et al.  (2019) 112 CTE >18% PMN 21 DPP Plasma protein levels Nyabinwa  et al.  (2020 b ) 436 † SCE ≥5% PMN Impact on reproductive performance Peter  et al.  (2018) 116 SCE  ≥5% ‡ Monitoring uterine health of cows with  Lactobacillus buchneri Plöntzke  et al.  (2010) 201 SCE ≥5% PMN Prevalence Salasel  et al.  (2010) 92 SCE ≥3% PMN Impact and risk factors Santos & Bicalho (2012) 16 EM – Bacteria colonies in uterine fluid Senosy  et al.  (2011) 198 SCE >5% PMN Metabolism investigated in SCE cows Sens & Heuwieser (2013) 149 SCE 4 to 18% PMN Bacterial infections’ effects on uterus Sharma  et al.  (2021) 41 SCE ≥5% PMN Back fat, serum leptin concentrations, endometrial cytology, milk yield Vallejo  et al.  (2018) 166 SCE >5% PMN Effect of EM on reproductive behaviour Vieira-Neto  et al.  (2014) 1569 CTE ≥5% PMN Reproductive performance effect of CTE and anovulation Yin  et al.  (2019) 3 – – MAPK pathway, LPS induction of inflammation and the effect of ferulic acid on EECs Zhao  et al.  (2018) 10 SCE – Activation of NF- κ B and promotion of miR-223 in LPS-induced EM *106 were subclinical;  † SCE = 13;  § BoHV-4 serum negative, healthy cows at slaughter;  ⁋ From cytobrush samples collected between 21 and 33 DPP or greater than 10% PMN between 34 and 47 DPP;  ‡ Cows were classified as having SCE. AI, artificial insemination; CM, chronic endometritis; CTE, cytological endometritis; DPP, days post partum; EECs, endometrial epithelial cells; EM, endometriosis; ESCs, endometrial stromal cells; PP, post partum; SCE, subclinical endometritis.\nFlow chart of pathophysiology themes and subthemes presented in the papers included in the study.\nOverview of included studies investigating the pathophysiology of subclinical endometritis in cows.\n*106 were subclinical;  † SCE = 13;  § BoHV-4 serum negative, healthy cows at slaughter;  ⁋ From cytobrush samples collected between 21 and 33 DPP or greater than 10% PMN between 34 and 47 DPP;  ‡ Cows were classified as having SCE.\nAI, artificial insemination; CM, chronic endometritis; CTE, cytological endometritis; DPP, days post partum; EECs, endometrial epithelial cells; EM, endometriosis; ESCs, endometrial stromal cells; PP, post partum; SCE, subclinical endometritis.\nA thematic analysis of the articles highlighted a crucial theme: the terminology and definition of endometritis. The most frequently used terms were SCE and cytological endometritis, used in 29 and 9 of the 44 articles, respectively ( Fig. 4 ). Although different terms were used, the diagnostic criteria were the same; the presence of PMN cells greater than 5% in a field of 300 endometrial cells, 35 days post partum and a lack of clinical signs and symptoms ( Wagener  et al.  2017 ).\nSCE has been commonly viewed to be the result of an underlying dysfunctional metabolic process and has been associated with metabolic risk factors ( Dubuc  et al.  2010  ,  Pascottini  et al.  2023 ). Depletion of energy stores has been postulated to increase the risk of SCE. A number of potential mechanisms have been proposed for this relationship within the literature ( Senosy  et al.  2011 ,  Sharma  et al.  2021 ). These stem from the relationship between glucose metabolism and the immune system. Low blood glucose levels in high yielding dairy cows are common due to the redirection of glucose for milk production. As a major fuel for immune cells, low blood glucose can lead to dysfunctional immune function. Blood glucose has been found to be lower in cows affected by SCE than control cows, reinforcing this link ( Senosy  et al.  2011 ). Furthermore, low back fat thickness (BFT) and body condition scores (BCS) have also been shown as risk factors for SCE. These have both been correlated with low blood glucose ( Sharma  et al.  2021 ), suggesting a link between SCE and metabolism-related risk.\nArtificial Insemination (AI) is extensively used in dairy herds for genetic improvement and economic benefits. It allows farmers to breed their cows with the best sires to produce optimal offspring. AI reduces the risk of infection transmitted through natural mating; however, if it is not carried out properly, there is the risk of introducing debris and bacteria into the cows’ uterus, which may trigger an immune response and subsequent endometritis, which is rare in the cow ( Parkinson & Morrell 2019 ).\nIn order to maintain optimum milk production, the aim is to get one calf, per cow, per year. With a gestation of around 285 days, a cow would need to be inseminated and become pregnant approximately 80 days after she last calved. This time period is called the calving to conception interval. To achieve a pregnancy by 80 days, a cow needs to be inseminated by around 60 days after calving, this is measured as the calving to first service interval. Health issues that occur in the first 60–80 days after calving can mean a cow is not suitable for insemination, therefore her days to first service, and calving to conception intervals will be prolonged. Both of these measures are an indicator of poor fertility, or poor reproductive health. In addition to metabolic dysfunction, peri- and post-partum risk factors for endometritis include calving related diseases such as retained placenta, dystocia, twin calving, and caesarean section ( Potter  et al.  2010 ,  Lee  et al.  2018 ).\nIt emerged in the thematic review of the bovine literature that the postpartum period is a high-risk period where cows are susceptible to SCE.\nA positive endometritis diagnosis correlates with a decrease in the total number of cows pregnant by 300 days post partum ( Gilbert  et al.  2005 ) and an increase in the calving to first service interval. Calving to conception interval and intercalving periods are significantly higher in cows who have had SCE in the postpartum period ( Barański  et al.  2013 ,  Sharma  et al.  2021 ). Cows who experience endometritis also need more inseminations to become pregnant (serves per conception) ( Gilbert  et al.  2005 ). Thus, the intercalving period is longer due to repeated failed AI attempts. In one study, 16.13% of cows who required more than one service per conception had previously suffered from SCE, whereas only 3.2% of cows pregnant to first service did ( Pascottini  et al.  2016 a ). In another study, the prevalence of cows who had SCE at AI was reported as 25.3% ( Pascottini  et al.  2017 ).\n Figure 4 Representation of the breakdown of terms used to describe non-‘clinical endometritis’ in the included studies.\nRepresentation of the breakdown of terms used to describe non-‘clinical endometritis’ in the included studies.\nAn important subtheme seen in a number of studies is infiltration of the endometrium with bacteria. Early studies in the cow of the causes of endometritis focused on isolation of microbiota from the uterus of diseased animals ( Williams  et al.  2005 ,  2007 ). One body of evidence suggests that endometritis, including SCE was caused by an infection of the endometrium, and several gram-negative bacteria were particularly correlated with disease.  Santos & Bicalho (2012)  found Fusobacteriaand Proteobacteria were the dominant phyla with Proteobacteria found in 70% of the study population.  Brodzki  et al.  (2014)  isolated both  E. coli  and  T. pyogenes  in affected groups and  Williams  et al.  (2005)  defined  E.coli, T.pyogenes , Fusobacteria and  Prevotella  as ‘uterine pathogens’ due to their consistent isolation from cows with uterine disease. While a number of bacterial species were isolated across different studies,  E. coli  was the most common ( Williams  et al.  2005 ,  2007 ,  Fábián  et al.  2008 ). α-Hemolytic streptococci (AHS) were also isolated in a study conducted by  Sens & Heuwieser (2013)  in the early postpartum period in cows with a previous diagnosis of SCE. AHS isolation was associated with an increase in PMN cell percentage and a higher risk of SCE at 4 weeks post-partum.\nIn recent years, the advancement of culture-independent molecular techniques has provided deeper insight into the microbiome of the bovine uterus, and its relationship with endometritis ( Santos & Bicalho 2012 ,  Miranda-CasoLuengo  et al.  2019 ). Temporal analysis has shown changes in the composition of cows with uterine disease from calving until late postpartum ( Santos & Bicalho 2012 ,  Knudsen  et al.  2016 ). Loss of bacterial diversity and dominance of the microbiome by a few bacterial taxa were related to the development of clinical disease ( Miranda-CasaLuengo  et al.  2019 ), and include the common uterine pathogens identified in culture-dependent studies. Studies, however, have reported a characteristic uterine microbiome in cows with cytological endometritis that is largely lacking the common uterine pathogens, and is instead dominated by commensal bacteria, suggesting that they may have a role in inflammatory processes in the endometrium ( Wang  et al.  2018 ). In contrast, other studies have found no difference in the microbiome between healthy cows and cows with SCE ( Pascottini  et al.  2020 ). A physiological response to bacterial contamination of the endometrium was proposed by  Vieira-Neto  et al.  (2014) . The intimate anatomic relationship between the ovarian arteries and uterine veins that is indispensable for the local transfer of prostaglandin from the endometrium to the ovary may also leak bacterial toxins to the developing follicle. These bacterial toxins disrupt the normal secretion of luteinising hormone (LH) and in doing so reduce the size and growth of the first postpartum dominant follicle, therefore impairing its ovulatory capacity. An increase in immune cell infiltration into the surface epithelium, lumen of endometrial glands and surrounding tissue causes occlusion and dilatation of glands. This leads to scar tissue formation in the uterus which is thought to impair implantation of the embryo and maintenance of pregnancy ( Vieira-Neto  et al.  2014 ).\nAn additional subtheme identified in the literature is that of viral infection, specifically bovine herpesvirus 4 (BoHV-4).  Fábián  et al.  (2008)  isolated BoHV-4 in 27/31 SCE cows. One cow had high viral loads in the absence of bacterial infection, however, in the majority of cases pathogenic bacteria were also isolated, suggesting BoHV-4 may not be a lone infectious agent in SCE.  Banks  et al.  (2008)  have suggested that viruses such as BoHV-4 may lie latent in white blood cells and are reactivated due to peripartum stress and endocrine conditions. It has also been suggested that BoHV-4 may have a positive tropism for endometrial cells ( Donofrio  et al.  2007 ,  Pascottini  et al.  2016 b \n ).\nA final subtheme that emerged from our analysis is the cellular pathways that underpin the uterine immune response. These involve the activation of several pro-inflammatory pathways following activation of toll-like receptors on the endometrial cell surface, which trigger downstream inflammatory components including NF -κ B, MAPK and the inflammasome. These are summarised in  Fig. 5 .\n Figure 5 Biomolecular pathways implicated in subclinical endometritis. (A) Metabolic stress: High systemic concentrations of non-esterified fatty acids (NEFAs), β-hydroxybutyrate (BHB), and acute phase proteins (APPs), along with dyscalcemia, contribute to neutrophil dysfunction and persistent endometrial inflammation. NEFAs may additionally activate toll-like receptor 4 (TLR-4). (B) Pathogenic pathway: Activation of TLR-4 by lipopolysaccharide (LPS) triggers: 1) PTEN pathway, promoting cell apoptosis by inhibiting the PI3K pathway; 2) MAP3K cascade, inducing AP-1, a transcription factor linked to apoptosis; 3) NF -κ B pathway, leading to pro-inflammatory cytokine release, neutrophil activation via inflammasome formation, and CXCR1/2 activation mediated by CXCL5 and IL-8 ( Peter  et al.  2018 ,  Fagundes  et al.  2019 ,  Yin  et al.  2019 ,  Jiang  et al.  2021 ,  Pascottini  et al.  2023 ).\nBiomolecular pathways implicated in subclinical endometritis. (A) Metabolic stress: High systemic concentrations of non-esterified fatty acids (NEFAs), β-hydroxybutyrate (BHB), and acute phase proteins (APPs), along with dyscalcemia, contribute to neutrophil dysfunction and persistent endometrial inflammation. NEFAs may additionally activate toll-like receptor 4 (TLR-4). (B) Pathogenic pathway: Activation of TLR-4 by lipopolysaccharide (LPS) triggers: 1) PTEN pathway, promoting cell apoptosis by inhibiting the PI3K pathway; 2) MAP3K cascade, inducing AP-1, a transcription factor linked to apoptosis; 3) NF -κ B pathway, leading to pro-inflammatory cytokine release, neutrophil activation via inflammasome formation, and CXCR1/2 activation mediated by CXCL5 and IL-8 ( Peter  et al.  2018 ,  Fagundes  et al.  2019 ,  Yin  et al.  2019 ,  Jiang  et al.  2021 ,  Pascottini  et al.  2023 ).\n\nThis systematic review with thematic analysis of 51 studies investigated the themes surrounding diagnostic methods and pathophysiology of SCE in dairy cows with the aim of identifying under-explored areas in the human condition, CE. The key themes that emerged within the literature were diagnosis and the pathophysiology of SCE. Within these themes, several subthemes were identified including, the significance of endometrial immune cells in aiding diagnosis and the influence of the uterine bacteria in the sequalae of disease. These themes present future potential avenues for research exploration in humans.\nA key theme that emerged was the use of cytology as a diagnostic tool. Cytology for CE has not been thoroughly explored in human literature; the current gold standard for diagnosis in humans is histological biopsy ( Hartman  et al.  2011 ,  Abdelazim  et al.  2015 ). Cytological endometrial sampling has been shown to reduce patient discomfort in humans in comparison to endometrial biopsy and thus may be a less invasive diagnostic approach for CE in humans ( Williams  et al.  2008 ,  Kumar 2017 ).\nThis present study highlights the importance of analysis of immune cells including PMN cells for the diagnosis of SCE in cows. PMN cells are activated in SCE in cows through a number of cellular pathways and result in further activation of neutrophils and other leukocytes ( Peter  et al.  2018 ,  Fagundes  et al.  2019 ,  Yin  et al.  2019 ). Diagnostic criteria for SCE centres on determining the proportion of PMN cells within the endometrial cell population, with specified levels indicating SCE at different time points after calving ( Kasimanickam  et al.  2004 ). Currently, there is no consensus regarding the diagnostic criteria for CE in humans ( Liu  et al.  2018 ) and cellular analysis does not focus on PMN numbers. The majority of human studies diagnose CE based on the quantification of plasma cells within the endometrial stroma ( Odendaal & Quenby 2021 ). The lack of a consensus regarding the diagnostic criteria, however, affects the prognostic ability of the test, impeding research in the area ( Liu  et al.  2018 ). The analysis of alternative immune cell populations, including PMN cells, in the human endometrium offers the potential for a diagnostic alternative or adjunct to current practice.\nAs well as diagnosis of SCE, pathophysiology emerged as a key theme in the present study. Subthemes of importance included the role of the endometrial microbiome in influencing disease, the cellular response to bacteria, and the impact of AI.\nThe isolation of gram negative bacteria from the uterus of cows that go on to receive a diagnosis of SCE was a subtheme explored across several of the studies. The predominant bacteria associated with disease were the ‘uterine pathogens’  E coli, T pyogenes ,  Fusobacteria  and  Prevotella  ( Gilbert  et al.  2005 ,  Williams  et al.  2005 ), with proposed bacterial sources including  Fusobacteria, Proteobacteria ,  E. coli , and AHS. Isolation of AHS was associated with an increased PMN cell percentage and increased risk of SCE at 4 weeks post-partum in cows ( Sens & Heuwieser 2013 ). Gram negative bacteria, specifically  E. coli , has been implicated in the development of CE in humans.  Cicinelli  et al.  (2009)  isolated  E. coli  in 11% of women with human CE in an infertility cohort. Indeed, the cell wall endotoxin produced by  E coli , lipopolysaccharide (LPS) is known to induce the upregulation of plasma cells following infection with  E. coli  in humans ( Kimura  et al.  2019 ,  Sola-Leyva  et al.  2021 ). In the cow, LPS is known to induce endometrial inflammation by binding to TLR4 on the surface of endometrial cells and initiating a downstream cascade of inflammatory mediators ( Jiang  et al.  2021 ). Thus, LPS induced inflammation is likely implicated in human CE but this remains to be explored.\nPotentiation of the development of SCE in cows by viral BoHV-4 emerged as a subtheme within the literature. It is suggested that BoHV-4 lies latent in white blood cells and upon changes in the endometrial environment and metabolic status of the cow, BoHV-4 reactivates ( Donofrio  et al.  2010 ). Investigation of the role of viruses in human CE is lacking, therefore the present study suggests this is an area that merits exploration.\nCellular pathways, including NF- κ B and MAPK, are a significant subtheme that has been implicated in the pathophysiology of SCE in cows. miRNA (miR-92b and miR-223) activation from these pathways suggests evidence of chronic disease ( Tahamtan  et al.  2018 ). The activation and effect of these pathways are underexplored within the human CE literature. Previous work has demonstrated their importance in colitis-induced colorectal cancer and inflammatory bowel disease, both chronic inflammatory conditions in humans ( Perera  et al.  2018 ,  Friedrich  et al.  2021 ). Both a decrease in miR-223 and subsequent increase in IL-1B and inflammasome activity are depicted as key players in the progression of these chronic conditions ( Mao  et al.  2018 ,  Dosh  et al.  2019 ). A similar pattern is seen in SCE in cows ( Zhao  et al.  2018 ). It is possible that the cellular pathways contributing to inflammation in the bovine endometrium are similar in women with CE. Indeed,  Di Pietro  et al.  (2018)  have already demonstrated differential expression of MiR-27a-3p and miR-124-3p in women with CE suggesting miRNA-mediated inflammatory pathways play a role in the pathogenesis of human CE. The interplay between miRNAs and markers of chronic inflammation may therefore act as both proxy markers for diagnosis of CE but may also point to underlying mechanistic pathways of disease.\nThe diagnosis of SCE in cows relies on identification of a high proportion of PMN cells within the uterus as determined by the relationship between PMN numbers and subfertility, or poor reproductive performance. Our thematic analysis identified poor reproductive performance as being associated with SCE in cows. In particular, cows which require more inseminations to conceive have a much higher prevalence of SCE than their herdmates which conceive upon their first insemination. This draws parallels with the impact of CE on human fertility ( Buzzaccarini  et al.  2020 ).\nIt is imperative to acknowledge certain limitations pertaining to the endometrial biology of the uterus in both the human as well as the bovine species. The endometrial shedding in women occurs cyclically at approximately 28-day intervals during menstruation (menstrual cycle). Conversely, reproductive cyclicity in cows manifests itself at approximately 21-day intervals, with the clinical expression being oestrus or heat, denoted as the oestrous cycle. Notably, cows do not undergo cyclic endometrial shedding, and the concept of menstruation is absent in this species. Furthermore, SCE in cows is predominantly observed in postpartum individuals facing challenges in adapting to the metabolic demands associated with heightened milk production. Metabolic stress, characterised by low glucose levels and elevated NEFAs, contributes to compromised innate immune function. This compromise includes the impaired ability of neutrophils to resolve inflammation and evacuate the uterine cavity, causing persistent endometrial inflammation in form of SCE. In contrast, in humans, CE is not inherently linked to the postpartum period or metabolic distress. The absence of such associations while underscoring points of mechanistic divergence also presents intriguing prospects for further research on CE among humans.\nThe present study is the first systematic review to explore diagnostic techniques and pathophysiology of SCE in cattle, using a thematic analysis, in order to highlight new areas of research for human CE. It is limited by the lack of direct comparisons between the condition in the two species to date.\nIn conclusion, our thematic systemic analysis of the literature on SCE in dairy cows has identified several areas that may be present within human CE yet remain underexplored. This offers the potential for new avenues of future research into human CE, and the possibility of novel, less invasive diagnostics and therapeutics for this disease.\n\nThe authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the study reported.\n\nThis research did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector. JO and NB’s time was funded by NIHR/MRC EME Project: 17/60/2. The open access fee for this article was funded by Tommy's.\n\nJO and SQ conceived of the project. KK and JO undertook the search, screening, data collection and data analysis. Initial manuscript preparation was conducted by KK and NB. KK, NB, EW, OBP, ST, SQ and JO reviewed and critically revised the manuscript.","source_license":"CC-BY-4.0","license_restricted":false}