Are Eosinophils a Specific Marker of Chronic Endometritis?

Introduction

Historically, endometritis has been classified as acute or chronic based on pathologic findings. The “gold standard” for the diagnosis of chronic endometritis is identification of plasma cells in the endometrial stroma, indicating persistent inflammation of the endometrium. ¹ Previous studies have shown that chronic endometritis is found in 3%–10% of women who undergo endometrial biopsy for abnormal uterine bleeding. ²

Patients with chronic endometritis are often asymptomatic or can present with vague symptoms, such as abnormal uterine bleeding and cramping lower-abdominal pain. Given that the clinical symptoms are neither specific nor sensitive, cases are potentially untreated, contributing to abnormal uterine bleeding and infertility. Recent studies have identified increased rates of chronic endometritis in patients experiencing recurrent pregnancy losses and recurrent implantation failures.^3,4

Diagnosis has limitations because of the difficulty in identifying plasma cells via biopsy. Previous studies have shown routine hematoxylin and eosin (H&E) staining does not allow easy discrimination between endometrial plasma cells and stromal cells.^5,6 The use of hysteroscopy for diagnosing chronic endometritis based on the identification of endometrial inflammation remains controversial. It has therefore been suggested to use immunohistochemical staining to facilitate visualization of plasma cells. ² However, this adds time and expense to the diagnostic process and is not used routinely. ⁷

More-recent studies have assessed the potential utility of a surrogate marker on routine staining for assessing chronic endometritis. Eosinophils—which comprise a type of chronic inflammatory cell and which are easily identified with H&E staining—have been suggested as a marker for chronic endometritis. Adegboyega et al. suggested that, while eosinophils may not be a diagnostic marker, their presence should raise suspicion of the diagnosis of chronic endometritis. ⁸ The current study was performed to evaluate how specific eosinophils are in identification of chronic endometritis.

Materials and Methods

This was a retrospective study of cases previously diagnosed as chronic endometritis by either endometrial biopsy or dilatation and curettage. The Rutgers Biomedical and Health Sciences institutional review board approved the study. The pathology departmental database at University Hospital in Newark, NJ, was searched for cases of chronic endometritis, which occurred from January 1, 1991 to August 1, 2015. Controls were cases of endometrial biopsies that were proliferative or showed disordered proliferation. Atrophy, pregnancy, secretory endometrium, and malignancy were exclusion criteria. Two viewers reviewed each randomly selected case and control to confirm the prior diagnosis and to evaluate for the presence or absence of eosinophils. Chronic endometritis was defined as identification of a single plasma cell. A minimum of 20 fields at 20 × magnification was reviewed per case. Differences in the number of eosinophils between the control group and the endometritis case group were assessed using Fisher's exact tests. A p-value of <0.05 was considered significant. Sensitivity, specificity, negative predicative value, and positive predictive value were all calculated.

Results

A total of 100 samples were reviewed; these included 50 cases and 50 controls. In the cases, eosinophils were present in 31 (62%) and absent in 19 (38%). In the controls, eosinophils were present in 19 (38%) and absent in 31 (62%; Table 1; p = 0.03). Sensitivity, specificity, positive predictive value, and negative predictive values were all 62%.

Of note, 5, cases that were initially diagnosed as proliferative endometrium with no mention of chronic endometritis were found to have plasma cells and were reassigned to the case group.

Discussion

Chronic endometritis is not a clinical diagnosis, as it is often asymptomatic, and, when symptoms are present, they are often vague, consisting of abnormal uterine bleeding, pelvic pain, leucorrhea, or dyspareunia. ⁹ Chronic endometritis can only be diagnosed by histopathologic examination, which can be challenging as plasma cells can be difficult to identify and resemble endometrial stromal cells, particularly when decidualized in the secretory phase. To circumvent this problem, cases of secretory endometrium were eliminated from the controls. The typical plasma cell has a “spoke-wheel” or “clock-face” pattern, but this morphology is not always seen. ⁹

It has also been shown that identification of plasma cells by conventional tissue staining can be hampered by their similar appearance to mononuclear leukocytes that reside in the human endometrium. ¹⁰ More-definitive diagnosis has been introduced with the use of immunohistochemistry. CD138, also known as syndecan-1, is a transmembrane-type heparan sulfate proteoglycan expressed on the plasma membranes of various types of somatic cells. Anti-CD138 antibodies have been shown to interact specifically with the CD138 antigens expressed on plasma cells. ¹¹ While studies have shown that immunohistochemistry for CD138 is superior for detection of plasma cells than H&E staining, use of the newer technique incurs additional costs ⁷ and can delay diagnosis. It is used in some cases but has not become the standard of care. As such, having an indicator to determine which cases required CD138 staining would be helpful.

The current study was conducted to determine if the presence of eosinophils—which are easier to recognize on routine staining, compared to plasma cells—could be used to make the diagnosis of chronic endometritis. It was found that, while the number of cases of chronic endometritis showing eosinophils was significantly greater than controls, 38% of controls also showed eosinophils. Thus, the presence of eosinophils is not sufficiently specific to be useful in making the diagnosis of endometritis.

Interestingly, 5 cases were found that were initially diagnosed as proliferative endometrium, with no signs of chronic endometritis, to have plasma cells upon further review. This showed how difficult the diagnosis is in routine practice, as high-power screening of endometrial stroma in the manner used to perform the current study is not utilized in diagnostic work, unless there are indicators of possible endometritis, with stromal alterations seen on low power, such as hypercellularity, spindled stroma, or areas of high- and low-stromal cellularity.

Conclusions

While eosinophils were more likely to be seen in chronic endometritis, they were frequently seen in controls as well. Although the number of cases of chronic endometritis with eosinophils was significantly greater than the number of controls with eosinophils, 38% of controls showed eosinophils, indicating that the presence of eosinophils is not a specific marker of chronic endometritis. The presence of eosinophils in the absence of identified plasma cells may suggest the need for CD 138 immunostaining.