Chronic endometritis diagnosis and fertility outcomes: an old unresolved question.

Abstract: Chronic endometritis, defined by chronic inflammation of the endometrium, remains a clinical and biologic challenge even using hysteroscopy allowing a direct vision of the uterine cavity without anesthesia, and conventional histology using Hematoxylin and Eosin staining. Our primary objectives were to evaluate the relevance of hysteroscopy and conventional histology compared to immunohistochemical expression of syndecan-1 (CD138, a marker of plasma cells), which is a heparan sulfate proteoglycan involved in inflammation and enables diagnosis of chronic endometritis. The second objective was to evaluate the impact of antibiotics on pregnancy rate. A retrospective study was conducted involving infertile women undergoing hysteroscopy and endometrial biopsy. Chronic endometritis was assessed using hysteroscopic findings and conventional histology compared to CD138 immunostaining. Effects of antibiotic therapy on CD138 expression on a second biopsy and on pregnancy rate were evaluated. Among the 661 infertile patients, 51 underwent hysteroscopy and endometrial biopsy. Twenty-three had a normal uterine cavity (45%) and among 28 patients with abnormal uterine cavity, ten (35.7%) had hysteroscopic findings of chronic endometritis. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of the hysteroscopy were 22, 100, 100, and 17%, with an infinite OR and an accuracy of 68.6, and 61.4, 100, 100%, 2.9%, and 66.7% respectively for conventional histology. The correlation coefficient between the first and second reading following CD138 immunostaining was moderate (Cohen's Kappa: 0.44 (95% CI: -0.059; 0.767) but good for plasma cell quantification (intraclass correlation coefficient 0.948). Plasma cell count was not predictive of pregnancy rate (P = 0.65) with an OR of 1.00. Pregnancy rate was significantly higher in treated patients (53%, 10/19) than in untreated patients (20%, 5/25) with an OR of 4.4 (95% CI: 1.17-16.8; P = 0.03).

Summary: Chronic endometritis is a reversible cause of infertility and remains a clinical and biologic challenge even using hysteroscopy and conventional histology and relies on the presence of plasma cells in immunohistochemistry. Our results underline the low accuracy of hysteroscopy and conventional histology to assess chronic endometritis, thus supporting the systematic use of CD138 immunostaining in infertile women even in the case of normal endometrium. Moreover, pregnancy rate seems enhanced by antibiotic therapy.