Outcome of assisted reproduction in women with congenital uterine anomalies: a prospective observational study.

OBJECTIVES: To assess the prevalence of congenital uterine anomalies, including arcuate uterus, and their effect on reproductive outcome in subfertile women undergoing assisted reproduction.

METHODS: Consecutive women referred for subfertility between May 2009 and November 2015 who underwent assisted reproduction were included in the study. As part of the initial assessment, each woman underwent three-dimensional transvaginal sonography. Uterine morphology was classified using the modified American Fertility Society (AFS) classification of congenital uterine anomalies proposed by Salim et al. If the external contour of the uterus was uniformly convex or had an indentation of < 10 mm, but there was a cavity indentation, it was defined as arcuate or septate. Arcuate uterus was further defined as the presence of a concave fundal indentation with a central point of indentation at an obtuse angle. Subseptate uterus was defined as the presence of a septum, not extending to the cervix, with the central point of the septum at an acute angle; if the septum extended to the internal cervical os, the uterus was defined as septate. Reproductive outcomes, including live birth, clinical pregnancy and preterm birth, were compared between women with a normal uterus and those with a congenital uterine anomaly. Subgroup analysis by type of uterine morphology and logistic regression analysis adjusted for age, body mass index, levels of anti-Müllerian hormone, antral follicle count and number and day of embryo transfer were performed.

RESULTS: A total of 2375 women were included in the study, of whom 1943 (81.8%) had a normal uterus and 432 (18.2%) had a congenital uterine anomaly. The most common anomalies were arcuate (n = 387 (16.3%)) and subseptate (n = 16 (0.7%)) uterus. The rate of live birth was similar between women with a uterine anomaly and those with a normal uterus (35% vs 37%; P = 0.47). The rates of clinical pregnancy, mode of delivery and sex of the newborn were also similar between the two groups. Preterm birth before 37 weeks' gestation was more common in women with uterine anomalies than in controls (22% vs 14%, respectively; P = 0.03). Subgroup analysis by type of anomaly showed no difference in the incidence of live birth and clinical pregnancy for women with an arcuate uterus, but indicated worse pregnancy outcome in women with other major anomalies (P = 0.042 and 0.048, respectively).

CONCLUSIONS: Congenital uterine anomalies as a whole, when defined using the modified AFS classification, do not affect clinical pregnancy or live-birth rates in women following assisted reproduction, but do increase the incidence of preterm birth. The presence of uterine abnormalities more severe than arcuate uterus significantly worsens all pregnancy outcomes. Copyright © 2017 ISUOG. Published by John Wiley & Sons Ltd.