Densities of decidual high endothelial venules correlate with T-cell influx in healthy pregnancies and idiopathic recurrent pregnancy losses

STUDY QUESTION: Do high endothelial venules (HEVs) appear in the uterus of healthy and pathological pregnancies? SUMMARY ANSWER: Our study reveals that HEVs are present in the non-pregnant endometrium and decidua parietalis (decP) but decline upon placentation in decidua basalis (decB) and are less abundant in decidual tissues from idiopathic, recurrent pregnancy losses (RPLs). WHAT IS KNOWN ALREADY: RPL is associated with a compromised decidual vascular phenotype. STUDY DESIGN, SIZE, DURATION: Endometrial (n = 29) and first trimester decidual (n = 86, 6-12th week of gestation) tissue samples obtained from endometrial biopsies or elective pregnancy terminations were used to determine the number of HEVs and T cells. In addition, quantification of HEVs and immune cells was performed in a cohort of decidual tissues from RPL (n = 25). PARTICIPANTS/MATERIALS, SETTING, METHODS: Position and frequency of HEVs were determined in non-pregnant endometrial as well as decidual tissue sections using immunofluorescence (IF) staining with antibodies against E-selectin, intercellular adhesion molecule, von Willebrand factor, ephrin receptor B4, CD34 and a carbohydrate epitope specific to HEVs (MECA-79). Immune cell distribution and characterization was determined by antibodies recognizing CD4S and CD3 by IF staining- and flow cytometry-based analyses. Antibodies against c-c motif chemokine ligand 21 (CCL2I) and lymphotoxin-beta were used in IF staining and Western blot analyses of decidual tissues. MAIN RESULTS AND THE ROLE OF CHANCE: Functional HEVs are found in high numbers in the secretory endometrium and decP but decline in numbers upon placentation in decB (P < 0.001). Decidua parietalis tissues contain higher levels of the HEV-maintaining factor lymphotoxin beta and decP-associated HEVs also express CCL2I (P <= 0.05), a potent T-cell chemoattractant Moreover, there is a positive correlation between the numbers of decidual HEVs and the abundance of CD3+ cells in decidual tissue sections (P <= 0.001). In-depth analysis of a RPL tissue collection revealed a decreased decB (P <= 0.01) and decP (P <= 0.01) HEV density as well as reduced numbers of T cells in decB (P <= 0.05) and decP (P <= .001) sections when compared with age-matched healthy control samples. Using receiver-operating characteristics analyses, we found significant predictive values for the ratios of CD3/CD45 (P < 0.001) and HEVs/total vessels (P < 0.001) for the occurrence of RPL. LIMITATIONS, REASONS FOR CAUTION: Analyses were performed in first trimester decidual tissues from elective terminations of pregnancy or non-pregnant endometrium samples from patients diagnosed with non-endometrial pathologies including cervical polyps, ovarian cysts and myomas. First trimester decidual tissues may include pregnancies which potentially would have developed placental disorders later in gestation. In addition, our cohort of non-pregnant endometrium may not reflect the endometrial vascular phenotype of healthy women. Finally, determination of immune cell distributions in the patient cohorts studied may be influenced by the different modes of tissue derivation. Pregnancy terminations were performed by surgical aspiration, endometrial tissues were obtained by biopsies and RPL tissues were collected after spontaneous loss of pregnancy. WIDER IMPLICATIONS OF THE FINDINGS: In this study, we propose an inherent mechanism by which the endometrium and in particular the decidua control T-cell recruitment. By demonstrating reduced HEV densities and numbers of T cells in decB and decP tissues of RPL samples we further support previous findings reporting an altered vascular phenotype in early pregnancy loss. Altogether, the findings provide important information to further decipher the etiologies of unexplained RPL.