Histopathological Association Between Chronic Endometritis and Adenomyosis: Clinical Findings and Risk Factors

This study was approved by the Ankara University Clinical Research Ethics Committee (Ethics No: i1‐03‐20). All procedures performed in this study were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki Declaration and its later amendments.

A total of 543 patients who underwent hysterectomy for benign gynecological conditions between January 2016 and December 2022 were initially assessed for eligibility. After applying the inclusion and exclusion criteria, 292 patients were included in the final analysis. PSM was used to create two balanced groups: 73 patients with adenomyosis group and 73 patients without adenomyosis (Figure  1 ). The demographic and clinical characteristics of the study population are summarized in Table  1 . The mean age of the patients was 48.8 ± 6 years, and the mean BMI was 28.7 ± 4.9 kg/m 2 . Patients in the adenomyosis group were significantly younger than those in the control group (47.1 ± 4.2 vs. 50.4 ± 7 p  = 0.012). No significant differences were observed between the groups in terms of BMI, parity, history of previous uterine surgery, or the presence of uterine fibroids ( p  > 0.05 for all). In the adenomyosis group, 17 (23.3%) patients and in the control group, 7 (9.6%) patients were diagnosed with CE, and a statistically significant difference was detected ( p  < 0.05). Demographic variables. Abbreviations: BMI, body mass index; CE, chronic endometritis; Hb, hemoglobin; HPF, high‐power field; Mean ± SD, number (percentages); NS, not significant. Plasma cells were quantified in five randomly selected stromal HPF per case, and the average number per HPF is reported. Patients with baseline endometrial loss were not included in the calculation. Basal endometrial thickness was measured in the endo‐myometrial junction zone in 112 (76.7%) patients, while basal endometrial loss was observed in 34 (23.3%) patients (Figure  2 ). Among the patients with basal endometrial loss, 16 (47%) were diagnosed with CE, whereas only 8 (7.1%) of the patients with measurable basal endometrial thickness had CE. This difference was statistically significant ( p  < 0.001). In patients whose basal endometrial thickness could be evaluated, the relationship between thickness and CE was also evaluated. A ROC curve analysis was performed to evaluate the relationship between basal endometrial thickness and CE. The optimal cutoff value for basal endometrial thickness in predicting CE was determined to be 0.15 mm, with a sensitivity of 83.3% and a specificity of 86.9% (area under the curve [AUC]: 0.888, 95% confidence interval [CI]: 0.798–0.977; p  < 0.001) (Figure  4 ). Receiver operating characteristic (ROC) curve evaluating the diagnostic performance of basal endometrial thickness for predicting chronic endometritis (CE). The ROC analysis was conducted in patients with measurable basal endometrial thickness to assess its ability to predict CE. The area under the curve was 0.888 (95% CI: 0.798–0.977; p  < 0.001), with a sensitivity of 83.3% and specificity of 86.9% at the optimal cutoff of 0.15 mm, as determined by the maximum Youden Index. The red diagonal line represents the line of no discrimination. Univariate logistic regression analysis revealed a statistically significant association between CE and adenomyosis (odds ratio [OR]: 2.89, 95% CI: 1.12–7.45; p  = 0.028) as well as between CE and basal endometrial loss (OR: 11.23, 95% CI: 4.56–27.64; p  < 0.001). However, in the multivariate logistic regression analysis, only basal endometrial loss remained an independent risk factor for CE (adjusted OR: 10.45, 95% CI: 4.12–26.51; p  < 0.001) (Table  2 ). Risk factors for Chronic Endometritis in Univariate and Multivariate Regression Analysis. Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio. We analyzed the relationship between BMI and CE risk, starting with BMI as a continuous variable. This analysis showed no statistically significant association with CE risk. However, when BMI was categorized into three groups ( 30 kg/m), which revealed a nonlinear association with CE, the results revealed that individuals with a BMI > 30 had a statistically significant association with CE risk. Its data were included in the analysis. Parameters with a p value below 0.1 in univariate analysis were included in multivariate analysis.

This study investigated the association between CE and adenomyosis using a propensity score‐matched case–control design. Our findings demonstrate a significant relationship between CE and adenomyosis, with CE prevalence being markedly higher in patients with adenomyosis (23.3% vs. 9.6%, p  < 0.05). Furthermore, basal endometrial loss emerged as a critical histopathological feature strongly associated with CE, independent of adenomyosis. These results align with emerging evidence suggesting that chronic inflammation may play a pivotal role in the pathogenesis of adenomyosis, particularly through mechanisms such as TIAR. Although basal endometrial loss is more commonly associated with intrauterine adhesions, emerging data suggest that it may also play a role in the pathogenesis of adenomyosis, possibly through shared mechanisms of tissue injury and repair. In our study, although basal endometrial loss was not independently associated with adenomyosis in multivariate analysis, its strong association with CE supports the hypothesis that chronic inflammation‐induced damage to the basal layer may be an early pathological event. Further research is needed to determine whether such alterations contribute causally to adenomyotic invasion or reflect a secondary process in CE‐related uterine remodeling. Many studies have noted a strong correlation between CE and endometriosis, with varying prevalence rates across different population groups. For instance, Cicinelli et al. reported a significantly higher incidence of CE in patients with endometriosis compared to those without (42.3% vs. 15.4%) [ 9 ]. Similarly, Takebayashi et al. documented a higher rate of concurrent CE among women diagnosed with endometriosis (36.6% vs. 9.4% in controls) [ 10 ]. Despite these associations, only limited research has explored the relationship between adenomyosis and CE. In a study conducted by Khan et al. in 2021, the researchers evaluated the relationship between CE and adenomyosis [ 17 ]. They found that 60% of the patients with diffuse adenomyosis and 58.8% of those with focal adenomyosis had CE. Additionally, in patients with focal adenomyosis, the presence of CE was significantly higher in the ipsilateral region compared to the contralateral side (58.8% vs. 11.7%). The high prevalence of CE may be attributed to the diagnostic criteria of the presence of one or more plasma cells, as well as the fact that all patients received GnRH analog treatment. In another recent study, endometrial samples from patients with and without adenomyosis were compared, revealing a 74% prevalence of CE in the adenomyosis group and 33.8% in the nonadenomyosis group [ 18 ]. Additionally, the study found a correlation between increased BMI and the presence of CE. Our study aimed to investigate the relationship between adenomyosis and CE. Our findings indicate a connection between these conditions. Unlike previous limited research, our study examined the potential association with endometrial thickness and loss of endometrial tissue, providing a more comprehensive understanding of their role in the pathogenesis. While we cannot claim a direct cause‐effect relationship between CE and adenomyosis, the pathogenesis of adenomyosis suggests that CE may play a role in the development of basal endometrium damage and inflammation through the TIAR mechanism. The TIAR mechanism refers to a cyclic process of inflammation‐induced epithelial–mesenchymal transition, neoangiogenesis, fibrosis, and aberrant stromal remodeling [ 3 , 19 ]. Repeated injury to the basal endometrium, as may occur in CE, can activate this cascade and promote myometrial infiltration of endometrial tissue, which is characteristic of adenomyosis. This proposed mechanism serves as a conceptual link between chronic inflammation and adenomyosis development. CE may occur due to intrauterine microbial colonization and/or tissue inflammatory reactions [ 11 , 20 , 21 , 22 ]. Several authors have proposed a link between endometriosis and chronic inflammation [ 9 , 23 ]. These studies hypothesized two distinct phases in the development of endometriosis. Intrauterine microorganisms may play a crucial role in the development of endometriosis. When microorganisms initially stimulate pathogen recognition receptors, it triggers the activation of proinflammatory pathways. Toll‐like receptors not only respond to molecular patterns associated with various external pathogens but also activate a wide range of endogenous pro‐inflammatory molecular pathways. This increased expression of pro‐inflammatory molecules then leads to Nuclear Transcription Factor‐κB‐dependent sterile inflammation as a subsequent process. Thus, after the first wave of Toll‐like receptor activation comes a second significant wave of sterile inflammation [ 23 ]. Sterile inflammation is believed to have a significant impact on the development of endometriosis and adenomyosis. Additionally, the unique uterine contraction patterns observed in women with CE further support the TIAR mechanism and the role of CE in adenomyosis [ 24 ]. Additionally, the presence of distinctly different microbiotic populations in the uterus of patients with endometriosis and adenomyosis, as compared to healthy individuals, further bolsters this theory [ 25 ]. Recent studies by various researchers have indicated that CE can interfere with both spontaneous and assisted reproductive technology‐induced pregnancies, leading to infertility, ART failures, abortion, and other obstetric complications [ 5 , 6 , 26 , 27 ]. Additionally, implantation failure and increased abortion rates observed in adenomyosis patients may be linked to the high prevalence of concomitant CE [ 6 , 28 , 29 ]. The notable presence of CE in adenomyotic patients as well as those with unexplained infertility and recurrent implantation failure suggests that antibiotic therapy could potentially have a positive impact on pregnancy outcomes [ 5 , 27 ]. There appears to be a clear relationship between CE and recurrent implantation failure and/or recurrent early pregnancy loss [ 30 ]. Similarly, patients with adenomyosis frequently experience implantation difficulties and adverse pregnancy outcomes. The shared clinical consequences of both CE and adenomyosis further corroborate our findings. This study has several limitations. First, its retrospective design introduces potential selection bias, although propensity score matching minimized confounding. Although patients with endometrial polyps and hydrosalpinx were excluded based on preoperative or histopathological findings, intrauterine adhesions were not systematically excluded, as no patients had a documented preoperative diagnosis or intraoperative suspicion of adhesions. Therefore, their presence cannot be completely ruled out and may represent a minor confounding factor in the interpretation of CE prevalence. Second, the sample size, though adequate for statistical analysis, may limit generalizability. Furthermore, as all samples were derived from hysterectomy specimens rather than timed hysteroscopic biopsies, we could not account for the phase of the menstrual cycle at the time of sampling. Additionally, information on recent hormonal treatments such as estrogen‐progestin or GnRH analogs was not systematically recorded. These factors may influence plasma cell density and endometrial morphology and represent a limitation of the current study. Larger, multicenter studies are needed to validate these findings. Third, the exclusion of mild adenomyosis cases (as only hysterectomy specimens were analyzed) may have skewed results toward more severe phenotypes. Another limitation of our study is that only participants with clearly measurable basal endometrial tissue were included in the ROC analysis, and endometrial thickness was evaluated as a continuous variable. Finally, the cross‐sectional nature of the study precludes causal inferences. An important methodological aspect of this study was the use of CD38 instead of CD138 for plasma cell identification. While CD138 is a well‐established marker, it has been shown to exhibit immunoreactivity in endometrial epithelial cells, which may lead to false‐positive interpretations, especially under suboptimal staining conditions [ 13 ]. CD38, in contrast, has been demonstrated to be more specific to stromal plasma cells with minimal epithelial background, offering a more accurate diagnostic approach when combined with conventional morphological criteria [ 13 , 14 , 15 , 16 ]. Our use of CD38 aimed to improve specificity and avoid overdiagnosis of CE, in line with recent recommendations [ 13 , 14 , 15 , 16 ]. Therefore, we believe that CD38 represents a reliable alternative in the context of retrospective hysterectomy specimens, and its use helped improve diagnostic accuracy and consistency in our study. This study provides robust evidence linking CE to adenomyosis, with basal endometrial loss serving as a key intermediary. These findings advance our understanding of adenomyosis pathogenesis and highlight CE as a potential therapeutic target. While further research is needed to unravel causal mechanisms, our results underscore the importance of integrating CE screening into the diagnostic workup of adenomyosis patients. Addressing chronic inflammation may not only improve symptom management but also disrupt the pathological cascade driving adenomyosis progression.

All authors consent to the publication of this manuscript. The manuscript does not contain any individual person's data in any form (including individual details, images, or videos).

Adenomyosis is a benign uterine disorder involving ectopic endometrial tissue within the myometrial wall. It often presents with clinical symptoms such as pelvic pain, heavy or irregular menstrual bleeding, and infertility [ 1 , 2 ]. Despite its significant impact on women's health, the exact etiology and pathogenesis of adenomyosis remain poorly understood. Current theories suggest that adenomyosis arises from the invagination of the basal endometrium into the myometrium, often accompanied by tissue injury and repair (TIAR) mechanisms, which may be influenced by chronic inflammation [ 3 , 4 ]. Chronic endometritis (CE), on the other hand, is a persistent inflammation of the endometrium, often caused by microbial colonization or immune dysregulation. Although CE is frequently asymptomatic or presents with nonspecific symptoms, it has been increasingly recognized as a contributing factor to various reproductive disorders, including recurrent implantation failure, recurrent pregnancy loss, and unexplained infertility [ 5 , 6 ]. The diagnosis of CE is challenging due to the lack of universally accepted diagnostic criteria, but the presence of plasma cells in endometrial tissue, often identified through immunohistochemical staining for markers such as CD38, is commonly used for its detection [ 7 , 8 ]. Recent studies have highlighted a potential link between chronic inflammation and the development of endometriosis, a condition that shares several pathogenic mechanisms with adenomyosis [ 9 , 10 ]. The archimetrosis theory posits that both endometriosis and adenomyosis originate from the basal endometrium, suggesting that chronic inflammation, as seen in CE, may play a role in their pathogenesis [ 3 ]. Specifically, the TIAR mechanism, which involves repeated tissue injury and repair due to inflammatory processes, has been proposed as a key factor in the development of adenomyosis [ 11 ]. However, the relationship between CE and adenomyosis has not been thoroughly investigated, and the existing literature on this topic remains limited. Upon reviewing the literature, it becomes apparent that the association between CE and adenomyosis has been explored to a limited degree compared to endometriosis. This study aims to investigate the prevalence of CE in patients with adenomyosis and to evaluate the potential role of CE in the pathogenesis of adenomyosis through a detailed histopathological and immunohistochemical analysis.

The authors declare no conflicts of interest.

This research employed a retrospective case–control design to explore the potential association between CE and adenomyosis. Histopathological samples were collected from patients who underwent hysterectomy for nonmalignant gynecological disorders at Ankara University school of Medicine, Department of Obstetrics and Gynecology, during the period from January 2016 to December 2022. The study protocol received ethical approval from the Ankara University Clinical Research Ethics Committee (approval number: i1‐03‐20). All procedures adhered to institutional and national research ethics standards. Women aged between 25 and 65 who underwent hysterectomy for benign reasons were considered for inclusion. The case group comprised individuals with histologically confirmed adenomyosis, while the control group included those without the condition. Subjects were excluded if they had received antibiotics or anti‐inflammatory treatments within the past 3 months, had current genital tract infections, or presented with preneoplastic or neoplastic lesions. Additionally, patients diagnosed with endometrial polyps or hydrosalpinx—based on preoperative evaluations or histological analysis—were excluded. Intrauterine adhesions were not systematically ruled out, as there was no consistent preoperative or clinical evidence indicating their presence. Patients with endometriosis or incomplete clinical documentation were also excluded from the study. Propensity score matching (PSM) was developed using a multivariate logistic regression model, which included factors such as age at surgery, body mass index (BMI), presence of fibroids, history of previous uterine surgery, parity, and use of tamoxifen [ 4 , 12 ]. Patients were matched 1:1 using nearest neighbor matching with a caliper width of 0.01. After matching, 73 patients with adenomyosis and 73 patients without adenomyosis were included in the final analysis (Figure  1 ). Flowchart of the patient selection process. Between January 2016 and December 2022, a total of 543 patients who underwent hysterectomy for benign gynecological conditions were screened. After applying exclusion criteria ( n  = 251), including recent antibiotic or anti‐inflammatory treatment, active genital infections, hydrosalpinx, endometrial polyps, neoplastic lesions, endometriosis, and age  65 years, 292 patients were eligible for analysis. Following 1:1 propensity score matching based on clinical covariates, 73 patients were assigned to the adenomyosis group and 73 to the control group. All endometrial tissue samples were obtained from hysterectomy specimens. As such, the exact timing of the menstrual cycle or recent hormonal treatments (e.g., low‐dose estrogen‐progestin or GnRH analogs) prior to surgery could not be consistently determined or controlled for in the analysis. All hysterectomy specimens were reevaluated by a senior gynecological pathologist (C.C.E.) who was blinded to the clinical data. The specimens were fixed in 10% formalin, embedded in paraffin, and sectioned at 4‐μm thickness. Hematoxylin and eosin (H&E) staining was performed to identify areas of adenomyosis and to measure the basal endometrial thickness in the endo‐myometrial junction zone. Basal endometrial thickness was measured in the endo‐myometrial junction zone using digital microscopy. Measurements were taken at three different points, and the average thickness was recorded. To ensure accurate vertical orientation and avoid measurement bias, only well‐oriented, full‐thickness regions of the endometrium and myometrium were selected for measurement. These three representative regions were chosen from H&E‐stained slides, which were scanned using a high‐resolution whole‐slide scanner (3DHISTECH Pannoramic 250 Flash3). Digital images were reviewed using the Case Viewer software, and quantitative morphometric measurements were performed using the Quant Center Application. The thickness was measured perpendicularly from the luminal surface to the basal endometrial border in each selected region, and the average of these three measurements was recorded. Basal endometrial loss was defined as the complete absence of the basal layer in at least two out of the three examined regions. If the basal layer was absent in only one region, the remaining two measurable fields were used to calculate the average thickness, and the case was not classified as having basal loss. This method ensured consistency and reduced observer bias. This method, demonstrated in Figure  2 , ensures standardization across cases. Representative hematoxylin and eosin (H&E)‐stained sections demonstrating basal endometrial thickness measurements at the endo‐myometrial junction. H&E‐stained full‐thickness sections were scanned using a digital whole‐slide scanner. Three well‐oriented, vertically cut regions were selected per case. Perpendicular measurements (indicated by black lines) were taken from the luminal surface to the endometrial‐myometrial interface at each of the three regions using Case Viewer and Quant Center software. The upper images show low‐power views used to select regions of interest, while the lower images show higher magnification with measurement lines and orientation. Scale bars are included. Basal endometrial loss was defined as the complete absence of the basal layer in at least two of the three measured regions. Plasma cells (PCs) were identified using immunohistochemistry with a monoclonal anti‐CD38 antibody. Although CD138 is widely used for plasma cell identification in CE, several studies have reported that CD138 can also be expressed on the basolateral membranes of endometrial epithelial cells, which may lead to interpretive challenges or potential overdiagnosis—particularly in suboptimal staining conditions [ 13 , 14 ]. CD38, in contrast, exhibits high specificity for stromal plasma cells and minimal reactivity in epithelial tissues. Therefore, to increase diagnostic accuracy and reduce potential misinterpretation, we used CD38 immunostaining, in combination with conventional nuclear morphology assessment (eccentric nuclei, “clockface” chromatin) as previously described in recent CE studies [ 15 , 16 ]. A senior gynecologic pathologist blinded to clinical data performed all evaluations. Tissue sections were deparaffinized, rehydrated, and subjected to antigen retrieval using a citrate buffer (pH 6.0). The sections were then incubated with a primary antibody against CD38 (clone SP149, Ventana Medical Systems, Tucson, AZ, USA) using an automated staining system (Ventana BenchMark XT, Roche Diagnostics, Basel, Switzerland). Positive staining was visualized using a diaminobenzidine chromogen, and the sections were counterstained with hematoxylin (Figure  3 ) [ 7 ]. CE was diagnosed based on the presence of ≥ 5 plasma cells per high‐power field (HPF) in the endometrial stroma, as previously described [ 7 , 8 ]. The number of plasma cells was counted in five different areas of the endometrial tissue, and the average count was recorded for descriptive analysis [ 7 , 8 ]. However, the diagnosis itself was based on the threshold criterion and not the mean count. Immunohistochemical staining of CD38 to identify plasma cells in chronic endometritis (CE). Representative serial sections showing CD38 immunoreactivity (brown membranous staining) in plasma cells located within the endometrial stroma. Top row: Low‐ and intermediate‐magnification views demonstrating overall distribution of plasma cells within the endometrial tissue. Bottom row: High‐power fields highlighting CD38‐positive plasma cells with distinct membranous staining (brown), aiding identification against a light blue hematoxylin background. Plasma cells were counted only when localized in the stromal compartment and showing characteristic morphology (eccentric nucleus, clockface chromatin). Epithelial staining was excluded from interpretation. Staining was performed using Ventana BenchMark XT automated system with anti‐CD38 antibody (clone SP149), and slides were counterstained with hematoxylin. Statistical evaluations were conducted using SPSS software version 26.0 (IBM Corp., Armonk, NY, USA) and R (version 4.0.4). Continuous data were reported as mean values alongside their standard deviations, while categorical variables were described in terms of frequencies and percentages. The Shapiro–Wilk test was applied to examine the normality of data distribution. Group comparisons were made using appropriate tests: either the Student's t ‐test or Mann–Whitney U test for continuous variables, depending on the distribution, and Pearson's chi‐square test or Fisher's exact test for categorical variables. To identify independent predictors of CE, both univariate and multivariate logistic regression models were employed. Variables that reached a p value threshold of < 0.1 in univariate analyses were included in the multivariate model. BMI was first analyzed as a continuous variable, but no significant association with CE was found. Therefore, BMI was categorized into three groups:  30 kg/m 2 . Patients with a BMI > 30 had a higher risk of CE, indicating a nonlinear relationship. Thus, the categorized variable was retained in the final model. To mitigate multicollinearity concerns, basal endometrial thickness and basal endometrial loss were not entered into the same multivariate regression model. Due to the inverse correlation between the two, only basal endometrial loss—defined as the total absence of the basal layer—was included, as it yielded clearer and more robust associations. Receiver operating characteristic (ROC) analysis was then carried out to assess the ability of basal endometrial thickness to predict CE. Only cases in which thickness could be measured were included in this analysis. To evaluate the predictive ability of basal endometrial thickness for the diagnosis of CE, a ROC curve analysis was performed using SPSS version 26. Only participants with clearly measurable basal endometrial tissue were included in this assessment. In this analysis, endometrial thickness was treated as a continuous variable. The threshold that provided the best balance between sensitivity and specificity was determined using the maximum Youden Index, which identified 0.15 mm as the optimal cutoff point. Participants in whom basal thickness could not be assessed in at least two out of three regions were excluded from the ROC analysis and analyzed separately. Statistical significance was defined as a p value < 0.05. In addition, PSM was carried out to control for potential confounding variables and ensure balanced comparison groups. The PSM algorithm relied on a multivariable logistic regression model incorporating the following covariates: age at the time of surgery, body mass index, presence of uterine fibroids, surgical history involving the uterus, parity, and tamoxifen usage. A 1:1 nearest‐neighbor matching approach was applied with a caliper distance set at 0.01. Trend score matching was implemented using R software to ensure statistical robustness.