Endometrial microbiota composition, but not chronic endometritis or hysteroscopy, predicts euploid miscarriage in unexplained RPL: associations with live birth and miscarriage outcomes.

Recurrent pregnancy loss (RPL) is a multifactorial condition, and reliable predictors of miscarriage remain limited. This observational cohort study investigated whether hysteroscopy, chronic endometritis (CE), and endometrial microbiome analysis can predict pregnancy outcomes in women with unexplained RPL. Ninety-nine patients with at least two previous miscarriages were enrolled, excluding those with antiphospholipid syndrome, chromosomal abnormalities in either partner, uterine anomalies, or a history of aneuploid miscarriage. CE was diagnosed by CD138-positive endometrial stromal cells, and affected patients received doxycycline 200 mg/day for 14 days. The primary outcome was cumulative live birth assessed in pregnancies resulting in either live birth or euploid miscarriage in relation to hysteroscopic findings, CE, and microbiome composition. The presence of Fusobacterium and Campylobacter was negatively associated with cumulative live birth. Focusing on Lactobacillus species, L. jensenii was positively associated with live birth. Additionally, the incidence of euploid miscarriage was significantly higher among cases with Lactobacillus abundance < 98% and the presence of Fusobacterium and Campylobacter. In contrast, CD138-positive cell counts and hysteroscopic findings were not predictive of pregnancy outcomes, and no associations were observed between the microbiome and hysteroscopic findings. However, the relative abundance of Lactobacillus tended to be lower when CD138-positive cell counts were higher. These findings suggest that Lactobacillus composition, as well as the presence of Fusobacterium and Campylobacter in the endometrial microbiome, may be associated with pregnancy outcomes, indicating that microbiome profiling could provide useful prognostic information for patients with unexplained RPL.