Risk of placenta praevia is linked to endometrial thickness in a retrospective cohort study of 4537 singleton assisted reproduction technology births

Abstract

STUDY QUESTION Is endometrial thickness measured prior to embryo transfer associated with placenta praevia? SUMMARY ANSWER Following IVF, the risk of placenta praevia is increased 4-fold in women with an endometrial thickness of >12 mm compared with women with an endometrial thickness of <9 mm. WHAT IS KNOWN ALREADY Placenta praevia is a serious complication of pregnancy with adverse maternal and neonatal outcomes. Placenta praevia is 2- to 6-fold more likely to occur following IVF treatment but it remains unknown what factors contribute to that increased risk. STUDY DESIGN, SIZE, DURATION Retrospective cohort study involving 4007 women who had 4537 singleton assisted reproduction technology (ART) births occurring between January 2006 and June 2012 with no loss to follow-up. The primary outcome measure was the diagnosis of placenta praevia, made by the treating obstetrician on a transvaginal ultrasound in the third trimester. PARTICIPANTS/MATERIALS, SETTING, METHODS Women who had singleton births following single embryo transfer performed at Monash IVF in Melbourne, Australia were included. Of the 4537 cycles leading to a singleton ART birth, 2951 were stimulated cycles with fresh embryo transfers; 355 were hormone replacement therapy frozen embryo transfers and 1231 were natural cycles with frozen embryo transfers. The dataset was analysed using binary logistic general estimating equations to calculate odds ratios for placenta praevia adjusted (aOR) for known confounders. MAIN RESULTS AND THE ROLE OF CHANCE The study groups did not differ significantly in age, BMI and aetiologies of infertility prior to IVF treatment. When compared with stimulated cycles, placenta praevia was less common in women undergoing natural cycles with frozen embryo transfers (OR 0.44, 95% confidence interval (CI) 0.27-0.70, P < 0.01) but hormone replacement therapy frozen embryo transfer cycles were not associated with a lower risk (OR 0.89, 95% CI 0.48-1.63). After adjusting for confounders, smoking (aOR 2.58, 95% CI 1.07-6.24, P = 0.04, endometriosis (aOR 2.01, 95% CI 1.21-3.33, P < 0.01) and endometrial thickness remained statistically significant as independent risk factors for placenta praevia. Compared with women with an endometrial thickness of <9 mm, women with an endometrial thickness of 9-12 mm had an aOR of 2.02 (95% CI 1.12-3.65, P = 0.02) and women with an endometrial thickness >12 mm had an aOR of 3.74 (95% CI 1.90-7.34, P < 0.01). These differences remained statistically significant after performing a sensitivity analysis limited to women with no previous births. LIMITATIONS, REASONS FOR CAUTION The study is retrospective in nature, not all confounders may have been accounted for and details on previous intrauterine surgery, a known risk factor, were not available. In addition, ultrasound assessments were carried out by several highly trained operators measuring the endometrial thickness, the main independent variable, in a two-dimensional plane and some inter-observer variability may therefore be present. WIDER IMPLICATIONS OF THE FINDINGS The findings of a higher risk of placenta praevia in patients with endometriosis and in those that smoke are in agreement with the current literature on natural conception. There have so far been no reports of an association between endometrial thickness and placenta praevia after ART. This novel finding warrants further study to elucidate the underlying cause of the association and to assess how to minimize harm to IVF patients and their offspring. The fact that the observed increased risk is not linked to the type of embryo transfer (fresh/frozen) but to the type of endometrial preparation, suggests that the risk of placenta praevia in ART can be reduced by considering an elective frozen embryo transfer in a natural cycle, especially given the growing evidence that this strategy also provides a number of other maternal and neonatal benefits.