Lower uterine segment scar assessment at 11-14 weeks' gestation to screen for placenta accreta spectrum in women with prior Cesarean delivery.

Abstract

To validate prospectively transvaginal ultrasound assessment of the lower uterine segment (LUS) scar at the time of first-trimester screening in women with previous Cesarean section (CS) and to determine its feasibility and accuracy in stratifying women according to the risk for placenta accreta spectrum (PAS) disorder. Women with a history of CS were recruited between 11 + 0 and 13 + 6 weeks' gestation and underwent LUS scar assessment using transvaginal ultrasound. A standardized midsagittal plane, which included the cervicoisthmic canal (CIC), the uterine scar and the placental site, was obtained. The scar was described in terms of its size (narrow or dehiscent) and its location in relation to the CIC (within or above), with each LUS scar classified into one of four groups based on these features. Placental location was assessed and classified as high- or low-lying. Women were stratified according to the risk of PAS, based on the relationship between the scar location and placental site. Women were considered high risk when the scar was above the CIC and the placenta was low-lying (i.e. when the placenta was overlying an exposed scar) and low risk when the scar was within the CIC and/or the placenta was high. High-risk patients were followed up at 20 weeks and 28-30 weeks for the development of PAS. Maternal demographics, detailed obstetric history and obstetric outcome were collected. First-trimester transvaginal ultrasound was offered to 535 women with prior CS during the study period. A LUS scar was visualized in 79.9% (401/502) of those who agreed to undergo the examination. At this scan, the LUS scar was above the CIC in 9.0% (36/401) of women, but only 5.7% (23/401)additionally had a low-lying placenta overlying the scar. Of these 23 high-risk women, two were found to have PAS on the mid-trimester screening scan and one was noted to have placental adherence during evacuation following mid-trimester termination of pregnancy. On the first-trimester scan, 94.3% (378/401) of women were at low risk of PAS. This screening protocol yielded a positive likelihood ratio of 21.33 (95% CI, 13.02-34.96), sensitivity of 100% (95% CI, 29.24-100%), specificity of 95.31% (95% CI, 92.39-97.35%), positive predictive value of 16.7% (95% CI, 5.8-39.2%) and negative predictive value of 100% (95% CI, 98.4-100%). On multivariable regression analysis performed to identify confounding variables associated with a LUS scar above the CIC, only maternal body mass index ≥ 30 kg/m2 was significant (odds ratio (OR), 2.42 (95% CI, 1.04-5.39); P = 0.03). Although there was a trend towards an increased risk of a LUS scar above the CIC in women with prior elective prelabor CS (OR, 1.72 (95% CI, 0.80-3.68)), this association did not reach statistical significance. Routine transvaginal ultrasound assessment of the location of the LUS scar and placenta at the time of first-trimester screening between 11 + 0 and 13 + 6 weeks' gestation in women with prior CS is a feasible and effective tool to identify those at risk of subsequent development of PAS disorder. A finding of placental implantation over an exposed LUS scar seems to be cardinal in predicting the risk of PAS disorder in women with prior CS, with an excellent negative predictive value. © 2021 International Society of Ultrasound in Obstetrics and Gynecology.